技术领域Technical field
本说明书中实施方式涉及车辆技术领域，特别是涉及一种车辆控制方法和自动驾驶提示方法及相关装置。The embodiments in this specification relate to the field of vehicle technology, and in particular, to a vehicle control method, an automatic driving prompting method, and related devices.
背景技术Background technique
车辆的自动驾驶已成为学术界与工业界研究与应用的重要方向。目前，自动驾驶技术的应用通常依赖于高精度导航地图。即，需要使用高精度导航地图实现车辆的高精度定位，以及进一步的运动规划和控制。Autonomous driving of vehicles has become an important direction for research and application in academia and industry. Currently, the application of autonomous driving technology usually relies on high-precision navigation maps. That is, high-precision navigation maps need to be used to achieve high-precision positioning of the vehicle, as well as further motion planning and control.
然而，高精度导航地图生产和更新的成本高难以快速适配人们活动范围内道路情况的动态变化。However, the high cost of producing and updating high-precision navigation maps makes it difficult to quickly adapt to dynamic changes in road conditions within the range of people's activities.
发明内容Contents of the invention
本说明书中多个实施方式提供一种车辆控制方法和自动驾驶提示方法及相关装置，可以无需使用高精度导航地图，实现自动驾驶。Multiple embodiments in this specification provide a vehicle control method, an automatic driving prompt method and related devices, which can realize automatic driving without using high-precision navigation maps.
本说明书的一个实施方式提供一种车辆控制方法，所述方法包括：在所述车辆从第一位置行驶至第二位置的过程中，获取定位数据和环境数据；其中，所述定位数据包括用于表示所述行驶过程中某一时刻的所述车辆的位置的信息，所述环境数据包括用于表示所述行驶过程中某一时刻的所述车辆的周围环境的信息；存储所述定位数据和所述环境数据；基于所述环境数据，获取与所述车辆的位置相关的交通标识信息；至少基于所述交通标识信息，获取第一自建地图信息；根据多于第一阈值的次数的行驶对应的第一自建地图信息，获取第二自建地图信息；其中，所述多于第一阈值的次数的行驶中的每一次行驶的路径包括至少一个相同的路段；其中，所述第二自建地图信息表示地图的精度高于所述第一自建地图信息；存储第二自建地图信息；至少基于所述第二自建地图信息，控制所述车辆在所述相同的路段执行自动驾驶。One embodiment of this specification provides a vehicle control method. The method includes: acquiring positioning data and environmental data while the vehicle is traveling from a first position to a second position; wherein the positioning data includes using Information representing the position of the vehicle at a certain moment during the driving process, the environmental data includes information representing the surrounding environment of the vehicle at a certain moment during the driving process; storing the positioning data and the environmental data; based on the environmental data, obtain traffic sign information related to the location of the vehicle; based at least on the traffic sign information, obtain first self-built map information; based on a number of times greater than the first threshold Travel the corresponding first self-built map information to obtain the second self-built map information; wherein the path of each trip in the number of trips exceeding the first threshold includes at least one identical road section; wherein the third The second self-built map information indicates that the accuracy of the map is higher than the first self-built map information; stores the second self-built map information; and controls the vehicle to execute on the same road section based on at least the second self-built map information. Autopilot.
本说明书的一个实施方式提供一种车辆的控制方法，应用于所述车辆，所述方法包括：确定所述车辆从指定的第一位置行驶至指定的第二位置涉及的路段；获取所述路段中至少部分子路段的自动驾驶信心信息；其中，所述自动驾驶信心信息为根据子路段对应的历史差异数据生成，所述历史差异数据是基于执行该子路段的自动驾驶决策算法生成的虚拟控制数据与驾驶员驾驶所述车辆驶过所述子路段执行的实际控制数据获得的；所述虚拟控制数据是由所述自动驾驶决策算法将根据自建地图信息、环境数据和定位数据进行处理，得到所述车辆的车道级定位的目标融合行驶轨迹，并基于所述目标融合行驶轨迹而生成的；其中，所述自建地图信息为所述车辆构建的地图信息；提示所述自动驾驶信心信息。One embodiment of the present specification provides a vehicle control method, which is applied to the vehicle. The method includes: determining a road segment involved in the vehicle traveling from a designated first position to a designated second position; obtaining the road segment The automatic driving confidence information of at least some sub-sections in the sub-section; wherein the automatic driving confidence information is generated according to the historical difference data corresponding to the sub-section, and the historical difference data is based on the virtual control generated by executing the automatic driving decision algorithm of the sub-section. The data is obtained from the actual control data performed by the driver driving the vehicle through the sub-section; the virtual control data is processed by the automatic driving decision-making algorithm based on self-built map information, environmental data and positioning data, Obtain the target fusion driving trajectory of the vehicle's lane-level positioning and generate it based on the target fusion driving trajectory; wherein the self-built map information is the map information constructed by the vehicle; prompt the automatic driving confidence information .
本说明书的一个实施方式提供一种车辆的控制装置，包括：第一获取模块，用于在所述车辆从第一位置行驶至第二位置的过程中，获取定位数据和环境数据；其中，所述定位数据包括用于表示所述行驶过程中某一时刻的所述车辆的位置的信息，所述环境数据包括用于表示所述行驶过程中某一时刻的所述车辆的周围环境的信息；存储所述定位数据和所述环境数据；第二获取模块，用于基于所述环境数据，获取与所述车辆的位置相关的交通标识信息；第三获取模块，用于至少基于所述交通标识信息，获取第一自建地图信息；第四获取模块，用于根据多于第一阈值的次数的行驶对应的第一自建地图信息，获取第二自建地图信息；其中，所述多于第一阈值的次数的行驶中的每一次行驶的路径包括至少一个相同的路段；其中，所述第二自建地图信息表示地图的精度高于所述第一自建地图信息；存储模块，用于存储第二自建地图信息；控制模块，用于至少基于所述第二自建地图信息，控制所述车辆在所述相同的路段执行自动驾驶。One embodiment of the present specification provides a vehicle control device, including: a first acquisition module, configured to acquire positioning data and environmental data while the vehicle is traveling from a first position to a second position; wherein, The positioning data includes information used to represent the position of the vehicle at a certain moment during the driving process, and the environmental data includes information used to represent the surrounding environment of the vehicle at a certain moment during the driving process; Store the positioning data and the environmental data; a second acquisition module, configured to acquire traffic sign information related to the location of the vehicle based on the environmental data; a third acquisition module, configured to obtain traffic sign information based on at least the traffic sign information to obtain the first self-built map information; the fourth acquisition module is used to obtain the second self-built map information according to the first self-built map information corresponding to the number of trips exceeding the first threshold; wherein, the more than The path of each trip in the first threshold number of trips includes at least one identical road section; wherein the second self-built map information indicates that the accuracy of the map is higher than the first self-built map information; the storage module uses and a control module configured to store the second self-built map information and control the vehicle to perform automatic driving on the same road section based on at least the second self-built map information.
本说明书的一个实施方式还提供一种自动驾驶提示装置，所述自动驾驶提示装置包括：确定模块，用于确定所述车辆从指定的第一位置行驶至指定的第二位置涉及的路段；信心信息获取模块，用于获取所述路段中至少部分子路段的自动驾驶信心信息；其中，所述自动驾驶信心信息为根据子路段对应的历史差异数据生成，所述历史差异数据是基于执行该子路段的自动驾驶决策算法生成的虚拟控制数据与驾驶员驾驶所述车辆驶过所述子路段执行的实际控制数据获得的；所述虚拟控制数据是由所述自动驾驶决策算法将根据自建地图信息、环境数据和定位数据进行处理，得到所述车辆的车道级定位的目标融合行驶轨迹，并基于所述目标融合行驶轨迹而生成的；其中，所述自建地图信息为所述车辆构建的地图信息；提示模块，用于提示所述自动驾驶信心信息。One embodiment of this specification also provides an automatic driving prompt device, which includes: a determination module for determining the road section involved in the vehicle traveling from the designated first position to the designated second position; confidence The information acquisition module is used to obtain the automatic driving confidence information of at least some sub-sections in the road section; wherein the automatic driving confidence information is generated based on the historical difference data corresponding to the sub-section, and the historical difference data is based on executing the sub-section. The virtual control data generated by the automatic driving decision-making algorithm of the road section is obtained from the actual control data performed by the driver driving the vehicle through the sub-road section; the virtual control data is obtained by the automatic driving decision-making algorithm based on the self-built map. Information, environmental data and positioning data are processed to obtain the target fusion driving trajectory of the vehicle's lane-level positioning, and is generated based on the target fusion driving trajectory; wherein the self-built map information is constructed by the vehicle Map information; prompt module, used to prompt the automatic driving confidence information.
本说明书的一个实施方式提供一种电子设备，所述电子设备包括存储器及处理器，所述存储器中存储有至少一条计算机程序，所述至少一条计算机程序由所述处理器加载并执行，以实现如上所述的车辆的控制方法，或者，实现如上的自动驾驶提示方法。One embodiment of the present specification provides an electronic device. The electronic device includes a memory and a processor. At least one computer program is stored in the memory. The at least one computer program is loaded and executed by the processor to implement The vehicle control method as described above, or the automatic driving prompting method as described above.
本说明书的一个实施方式提供一种计算机可读存储介质，所述计算机可读存储介质中存储有至少一条计算机程序，所述至少一条计算机程序被处理器执行时能够实现如上所述的车辆控制方法，或者，实现如上的自动驾驶提示方法。One embodiment of the present specification provides a computer-readable storage medium. The computer-readable storage medium stores at least one computer program. When the at least one computer program is executed by a processor, it can implement the vehicle control method as described above. , or implement the above automatic driving prompt method.
本说明书提供的多个实施方式，通过车辆行驶过程中，采集并存储所述车辆的定位数据和环境数据，进一步的，车辆使用存储的定位数据和环境数据建立指定道路的语义地图，如此，当所述车辆再次行驶在所述指定道路时，便可以基于所述指定道路的语义地图执行自动驾驶，从而实现在无高精度导航地图情况下的自动驾驶。In various embodiments provided in this specification, the vehicle's positioning data and environmental data are collected and stored while the vehicle is driving. Further, the vehicle uses the stored positioning data and environmental data to establish a semantic map of the designated road. In this way, when When the vehicle drives on the designated road again, it can perform automatic driving based on the semantic map of the designated road, thereby realizing automatic driving without a high-precision navigation map.
附图说明Description of the drawings
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without exerting creative efforts.
图1为本说明书的一个实施方式提供的一种车辆控制方法的流程示意图。Figure 1 is a schematic flowchart of a vehicle control method according to an embodiment of this specification.
图2为本说明书的一个实施方式提供的一种车辆控制方法的流程示意图。FIG. 2 is a schematic flowchart of a vehicle control method according to an embodiment of this specification.
图3为本说明书的一个实施方式提供的一种车辆控制方法的流程示意图。FIG. 3 is a schematic flowchart of a vehicle control method according to an embodiment of this specification.
图4为本说明书的一个实施方式提供的一种车辆控制方法的流程示意图。FIG. 4 is a schematic flowchart of a vehicle control method according to an embodiment of this specification.
图5为本说明书的一个实施方式提供的一种车辆控制装置的模块示意图。FIG. 5 is a schematic module diagram of a vehicle control device according to an embodiment of this specification.
图6为本说明书的一个实施方式提供的一种车辆控制方法的流程示意图。FIG. 6 is a schematic flowchart of a vehicle control method according to an embodiment of this specification.
图7为本说明书的一个实施方式提供的一种自动驾驶提示方法的流程示意图。Figure 7 is a schematic flowchart of an automatic driving prompting method provided by an embodiment of this specification.
图8为本说明书的一个实施方式提供的一种车辆控制装置的模块示意图。FIG. 8 is a schematic module diagram of a vehicle control device according to an embodiment of this specification.
图9为本说明书的一个实施方式提供的一种自动驾驶提示装置的模块示意图。Figure 9 is a schematic module diagram of an automatic driving prompt device provided in an embodiment of this specification.
图10为本说明书的一个实施方式提供的一种电子设备的示意框图。FIG. 10 is a schematic block diagram of an electronic device according to an embodiment of this specification.
具体实施方式Detailed ways
下面将结合本说明书中的附图，对本本说明书提供的实施方式中的技术方案进行清楚、完整地描述，显然，所描述的实施方式仅仅是一部分实施方式，而不是全部的实施方式。基于本说明书提供的实施方式，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。The technical solutions in the embodiments provided in this specification will be clearly and completely described below with reference to the accompanying drawings in this specification. Obviously, the described embodiments are only part of the embodiments, not all of them. Based on the embodiments provided in this specification, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.
高精度导航地图相对于普通导航地图是精度更高、数据维度更多的电子地图。精度更高可以体现在精确到厘米级别，数据维度更多体现在其包括了除道路信息之外的与交通相关的周围静态信息。Compared with ordinary navigation maps, high-precision navigation maps are electronic maps with higher precision and more data dimensions. The higher accuracy can be reflected in centimeter level accuracy, and the data dimension is more reflected in the fact that it includes surrounding static information related to traffic in addition to road information.
高精度导航地图将大量的行车辅助信息存储为结构化数据，这些信息可以分为两类。第一类是道路数据，比如车道线的位置、类型、宽度、坡度和曲率等车道信息。第二类是车道周边的固定对象信息，比如交通标志、交通信号灯等信息、车道限高、下水道口、障碍物及其他道路细节，还包括高架物体、防护栏、数目、道路边缘类型、路边地标等基础设施信息。在车辆执行自动驾驶的过程中，可以将依据高精度地图、车联网技术提供的多维度信息对具体驾驶问题做出判断、输出车辆控制信号并交给执行层执行。High-precision navigation maps store a large amount of driving assistance information as structured data, which can be divided into two categories. The first category is road data, such as lane information such as the position, type, width, slope and curvature of lane lines. The second category is fixed object information around the lane, such as traffic signs, traffic lights and other information, lane height limits, sewer openings, obstacles and other road details. It also includes overhead objects, guardrails, numbers, road edge types, roadside Infrastructure information such as landmarks. During the vehicle's autonomous driving process, specific driving issues can be judged based on multi-dimensional information provided by high-precision maps and Internet of Vehicles technology, and vehicle control signals can be output and handed over to the execution layer for execution.
高精度地图的实际应用可以包括制图过程、用图过程和更新过程三个紧耦合的过程，以保证数据的高频流动和更新。具体的，例如，制图过程可以包括外业采集和内业制作，用图过程可以包括高精（自）定位、环境感知和路径规划，更新过程可以包括变化检测和交叉验证等。The practical application of high-precision maps can include three tightly coupled processes: mapping process, map using process and updating process to ensure high-frequency flow and updating of data. Specifically, for example, the mapping process can include field collection and in-house production, the mapping process can include high-precision (self-) positioning, environment awareness, and path planning, and the updating process can include change detection and cross-validation.
高精度地图动态比较繁琐的制图过程，决定了后期的更新维护也会占据很大的工作量。车辆实现智能驾驶所需要的高精度地图依据更新频率可以划分为四类：更新频率为一个月的长期静态数据、更新频率为1小时的短期静态数据、更新频率为1分钟的半动态数据、更新频率为1秒钟的动态数据。与当前普及的普通导航地图1~2月更新一次的频率相比，高精度地图的更新频率之高、难度之大可想而知。高精度地图的制作和更新成本都非常高。The dynamic and cumbersome mapping process of high-precision maps determines that subsequent updates and maintenance will also take up a large workload. The high-precision maps required for vehicles to achieve intelligent driving can be divided into four categories based on update frequency: long-term static data with an update frequency of one month, short-term static data with an update frequency of 1 hour, semi-dynamic data with an update frequency of 1 minute, and Dynamic data with a frequency of 1 second. Compared with the current popular ordinary navigation maps, which are updated once every 1 to 2 months, it is conceivable that the update frequency of high-precision maps is high and difficult. High-precision maps are very expensive to produce and update.
在相关技术中，车辆实现自动驾驶，对于高精度导航地图有非常强的依赖性。因此，有必要提供一种车辆可以无需完全依赖高精度导航地图，实现自动驾驶的技术方案。In related technologies, vehicles that realize autonomous driving have a strong dependence on high-precision navigation maps. Therefore, it is necessary to provide a technical solution that enables vehicles to realize autonomous driving without completely relying on high-precision navigation maps.
本说明书的实施方式所描述的车辆可以是有人驾驶且有辅助智能驾驶功能的车辆，也可以是能够自动智能行驶车辆。车辆类型具体可以包括轿车、越野车、货车等，本说明书的实施方式对车辆不作具体限定。The vehicle described in the embodiments of this specification may be a human-driven vehicle with assisted intelligent driving functions, or may be a vehicle capable of autonomous intelligent driving. Vehicle types may specifically include cars, off-road vehicles, trucks, etc. The embodiments of this specification do not specifically limit the vehicles.
本说明书的一个实施方式提供一种车辆控制方法。所述车辆控制方法可以应用于车辆的车辆控制系统。所述车辆控制方法可以包括：熟路发现阶段、影子模式验证阶段、自动驾驶路径推荐阶段和自动驾驶阶段。所述熟路发现阶段可以实现控制车辆自动发现熟路，以及分析驾驶员的驾驶风格和道路状况分析。所述影子模式验证阶段可以通过对车辆的虚拟自动驾驶与驾驶员驾驶进行对比验证，得出可以实现自动驾驶的路径。所述自动驾驶路径推荐阶段可以实现针对能够自动驾驶的路径向驾驶员进行推荐，并与驾驶员进行交互确认。所述自动驾驶阶段可以实现在驾驶员允许的路径执行车辆的自动驾驶。One embodiment of this specification provides a vehicle control method. The vehicle control method may be applied to a vehicle control system of a vehicle. The vehicle control method may include: a familiar road discovery stage, a shadow mode verification stage, an automatic driving path recommendation stage, and an automatic driving stage. The familiar road discovery stage can control the vehicle to automatically discover familiar roads and analyze the driver's driving style and road conditions. The shadow mode verification stage can compare and verify the virtual automatic driving of the vehicle and the driver's driving to obtain a path that can realize automatic driving. The automatic driving path recommendation stage can recommend routes capable of automatic driving to the driver and conduct interactive confirmation with the driver. The automatic driving stage can realize automatic driving of the vehicle on a path allowed by the driver.
请参阅图1，在所述熟路发现阶段可以包括以下步骤。Referring to Figure 1, the familiar path discovery stage may include the following steps.
步骤S11：存储所述车辆行驶过程中，采集到的所述车辆的定位数据和环境数据。Step S11: Store the positioning data and environmental data of the vehicle collected during the driving of the vehicle.
在本实施方式中，所述定位数据可以用于确定车辆相对于地面的位置。如此，通过采集车辆的定位数据，便可以针对车辆所处的位置，进行连续的定位。在一些实施方式中，定位数据可以包括卫星定位数据。卫星定位数据可以用于表示车辆相对于地面的绝对定位。当然，定位数据还可以包括车载的惯导系统生成的定位数据。惯导定位数据用于表示车辆的相对定位。In this embodiment, the positioning data may be used to determine the position of the vehicle relative to the ground. In this way, by collecting the positioning data of the vehicle, continuous positioning can be performed based on the location of the vehicle. In some embodiments, positioning data may include satellite positioning data. Satellite positioning data can be used to represent the vehicle's absolute position relative to the ground. Of course, the positioning data may also include positioning data generated by the vehicle's inertial navigation system. Inertial navigation positioning data is used to represent the relative positioning of the vehicle.
在一些实施方式中，所述车辆控制系统可以接收所述车辆的卫星定位数据；其中，所述卫星定位数据具有第一频率；以及，获取所述车辆的惯导系统生成的惯导定位数据；其中，所述惯导定位数据具有第二频率；其中，所述第二频率大于所述第一频率；基于所述惯导定位数据校正所述卫星定位数据，得到所述车辆的定位数据；其中，所述车辆的定位数据具有所述第二频率。In some embodiments, the vehicle control system may receive satellite positioning data of the vehicle; wherein the satellite positioning data has a first frequency; and obtain inertial navigation positioning data generated by the inertial navigation system of the vehicle; Wherein, the inertial navigation positioning data has a second frequency; wherein the second frequency is greater than the first frequency; the satellite positioning data is corrected based on the inertial navigation positioning data to obtain the positioning data of the vehicle; wherein , the positioning data of the vehicle has the second frequency.
在本实施方式中，车辆通过卫星定位系统和接收器接收卫星定位数据。可以针对接收到的卫星信号进行解算，得到卫星定位数据。在一些实施方式中，可以采取多个模型针对卫星定位数据分别进行解算之后，选取最优解作为最终使用的卫星定位数据。In this embodiment, the vehicle receives satellite positioning data through a satellite positioning system and a receiver. The received satellite signals can be solved to obtain satellite positioning data. In some implementations, multiple models can be used to separately solve the satellite positioning data, and then the optimal solution is selected as the final satellite positioning data.
在本实施方式中，惯导系统是一个使用加速计和陀螺仪来测量物体的加速度和角速度，并可以连续估算运动物体位置、姿态和速度的辅助导航系统。通过检测惯导系统的加速度和角速度，惯导可以检测位置变化。具体的，例如，向东或向西的运动。速度变化，例如，速度大小或方向的变化。姿态变化，例如，绕各个轴的旋转。具体的，例如，可以通过针对陀螺仪输出信号的解算，得出表示姿态的姿态数据。结合速度的速度数据和姿态数据，便可以得出表示车辆的相对位置变化的位置变化数据。In this embodiment, the inertial navigation system is an auxiliary navigation system that uses accelerometers and gyroscopes to measure the acceleration and angular velocity of objects, and can continuously estimate the position, attitude, and speed of moving objects. By detecting the acceleration and angular velocity of the inertial navigation system, inertial navigation can detect changes in position. Specific, for example, eastward or westward movement. A change in velocity, for example, a change in magnitude or direction of velocity. Posture changes, for example, rotations about various axes. Specifically, for example, attitude data representing the attitude can be obtained by solving the gyroscope output signal. Combining the speed data and attitude data of the speed, position change data indicating the relative position change of the vehicle can be obtained.
在本实施方式中，惯导定位数据的第二频率高于卫星定位数据的第一频率。如此，为了提升针对车辆的位置的定位精度。可以采用高频率的惯导定位数据对相对低频率的卫星定位数据进行校正，提升卫星定位数据的频率。具体的，例如，可以先按照采集时间将卫星定位数据和惯导定位数据进行数据对齐，此时，由于卫星定位数据的数据频率小于惯导定位数据的数据频率，使得，部分惯导定位数据没有对应的卫星定位数据。如此，可以基于连续的惯导定位数据之间表示的位置变化数据，推断补全惯导定位数据对应的卫星定位数据。如此，使得补全后的卫星定位数据可以具有第二频率。In this embodiment, the second frequency of the inertial navigation positioning data is higher than the first frequency of the satellite positioning data. In this way, in order to improve the positioning accuracy of the vehicle's position. High-frequency inertial navigation positioning data can be used to correct relatively low-frequency satellite positioning data to increase the frequency of satellite positioning data. Specifically, for example, the satellite positioning data and the inertial navigation positioning data can be aligned according to the collection time. At this time, because the data frequency of the satellite positioning data is smaller than the data frequency of the inertial navigation positioning data, some of the inertial navigation positioning data does not Corresponding satellite positioning data. In this way, the satellite positioning data corresponding to the complementary inertial navigation positioning data can be inferred based on the position change data represented between the consecutive inertial navigation positioning data. In this way, the completed satellite positioning data can have the second frequency.
在本实施方式中，所述环境数据可以用于表示车辆行驶过程中，通过采集到的包括表示交通标识的传感数据。具体的，例如，车辆可以设置有多个检测设备和传感器，并可以通过该多个检测设备和传感器采集环境数据。具体的，例如，环境数据可以包括鸟瞰图数据或平面图数据。In this embodiment, the environmental data may be used to represent sensor data collected during driving of the vehicle including traffic signs. Specifically, for example, a vehicle may be equipped with multiple detection devices and sensors, and environmental data may be collected through the multiple detection devices and sensors. Specifically, for example, the environment data may include bird's-eye view data or plan view data.
在本实施方式中，定位数据和环境数据都可以分别对应有采集时间。如此，便可以通过采集时间将定位数据和环境数据相对应。具体的，可以理解为，通过采集时间为定位数据和环境数据建立对应关系，存在对应关系的定位数据和环境数据中，环境数据可以表示定位数据所表示的位置的环境信息。在本实施方式中，存储定位数据和环境数据，可以分别按照采集时间独立存储入数据库。在一些实施方式中，也可以通过采集时间为定位数据和环境数据建立对应关系之后，对应存储入数据库中。In this implementation, both positioning data and environmental data may have corresponding collection times respectively. In this way, positioning data and environmental data can be correlated through acquisition time. Specifically, it can be understood that a corresponding relationship is established between the positioning data and the environmental data through the collection time. Among the positioning data and the environmental data that have a corresponding relationship, the environmental data can represent the environmental information of the location represented by the positioning data. In this implementation, the positioning data and environmental data are stored and can be independently stored in the database according to the collection time. In some implementations, the positioning data and the environment data can also be stored in a database after establishing a corresponding relationship based on the collection time.
步骤S12：从存储的所述环境数据识别出对应所述定位数据的交通标识信息，形成单程建图信息。Step S12: Identify traffic sign information corresponding to the positioning data from the stored environmental data to form one-way mapping information.
在本实施方式中，交通标识信息可以用于表示交通规则。进一步的，交通标识信息表示车辆在具有该交通标识信息的道路上需要遵守该交通规则。如此，通过识别出交通标识信息，使得车辆在执行自动驾驶的过程中的运动规划可以遵循交通标识信息表示的交通规则。具体的，例如，识别出表示可以左转的交通标识信息，可以表示车辆可以在该交通路口左转。识别出地面上的表示直行车道的交通标识信息，表示车辆行驶在该车道时在交通路口不允许拐弯。In this embodiment, traffic sign information may be used to represent traffic rules. Further, the traffic sign information indicates that the vehicle needs to abide by the traffic rules on the road with the traffic sign information. In this way, by identifying the traffic sign information, the vehicle's motion planning during autonomous driving can follow the traffic rules represented by the traffic sign information. Specifically, for example, identifying traffic sign information indicating that a left turn is possible can indicate that the vehicle can turn left at the traffic intersection. The traffic sign information indicating the through lane on the ground is recognized, indicating that vehicles are not allowed to turn at the traffic intersection when driving in this lane.
在本实施方式中，不同定位数据对应的道路可以具有不同的交通标识信息。如此，交通标识信息本身表示的交通标识与位置存在对应关系。具体的，例如，一条道路会经过学校的校门口，道路上设置有表示减速的交通标识。需要从环境数据中对应定位数据识别出表示该交通标识的交通标识信息，如此，车辆在自动驾驶过程中途经该定位数据表示的位置时，可以依照该交通标识信息执行控制车速。具体的，例如，在车辆中可以设置有训练后的机器学习模型，通过将环境数据输入该机器学习模型识别出交通标识信息。交通标识信息可以包括地面交通标识信息和空中交通标识信息。地面交通标识可以包括车道线、停止线等。空中交通标识识别可以包括红绿灯、交通牌等识别。在一些实施方式中，针对红绿灯和交通牌进行识别的过程中，可以基于3D检测结果，与车辆的相机的外参关系，将红绿灯和交通牌的各角点分别投影到车辆的前视相机的PV（Perspective View）视角图像上。可以分别针对红绿灯和交通牌设置矩形框。如此，可以对红绿灯的矩形框中的图像进行红灯、黄灯和绿灯识别，对交通牌的矩形框中的图像进行交通标识的识别。In this embodiment, roads corresponding to different positioning data may have different traffic sign information. In this way, there is a corresponding relationship between the traffic sign represented by the traffic sign information itself and the location. Specifically, for example, a road passes through the gate of a school, and there are traffic signs indicating slowdown on the road. The traffic sign information representing the traffic sign needs to be identified from the corresponding positioning data in the environmental data. In this way, when the vehicle passes the location represented by the positioning data during autonomous driving, the vehicle speed can be controlled according to the traffic sign information. Specifically, for example, a trained machine learning model can be installed in the vehicle, and the traffic sign information can be recognized by inputting environmental data into the machine learning model. Traffic identification information may include ground traffic identification information and air traffic identification information. Ground traffic signs can include lane lines, stop lines, etc. Air traffic sign recognition can include the recognition of traffic lights, traffic signs, etc. In some embodiments, during the process of identifying traffic lights and traffic signs, each corner point of the traffic lights and traffic signs can be projected onto the vehicle's front-view camera based on the 3D detection results and the external parameter relationship with the vehicle's camera. PV (Perspective View) perspective image. Rectangular frames can be set separately for traffic lights and traffic signs. In this way, the image in the rectangular frame of the traffic light can be recognized as red light, yellow light and green light, and the image in the rectangular frame of the traffic sign can be recognized as a traffic sign.
在一些实施方式中，具体的，可以基于BEVFormer作为主干网，并针对不同交通标识设置相关的检测头，来实现得出表示障碍物和交通标识的环境数据。可以，基于一定的交通标识数据针对BEVFormer的检测头进行训练，提升BEVFormer的检测精度。当然，在一些实施方式中，也可以通过HDMapNet或者VectorMapNet实现。In some implementations, specifically, BEVFormer can be used as the backbone network, and relevant detection heads can be set for different traffic signs to obtain environmental data representing obstacles and traffic signs. Yes, the detection head of BEVFormer can be trained based on certain traffic sign data to improve the detection accuracy of BEVFormer. Of course, in some implementations, it can also be implemented through HDMapNet or VectorMapNet.
当然，在一些实施方式中，环境数据中还可以包括有表示道路中动态障碍物和静态障碍物的障碍物数据。可以在生成单程建图信息过程中，仅仅关注环境数据中的表示交通标识的交通标识数据，而过滤掉表示障碍物的障碍物数据。动态障碍物可以包括机动车、行人或非机动车等。静态障碍物可以包括水马、锥桶等。作为联合地图数据也可以仅仅过滤掉动态障碍物数据，保留静态障碍物数据。因此，动态障碍物数据不具有通用性，使得在生成联合地图数据时，可以不必保留。针对静态障碍物数据，可能会持续一定时间，可以随着车辆多次行驶于相应道路，随着实际道路中静态障碍物的情况，更新存储。Of course, in some implementations, the environment data may also include obstacle data representing dynamic obstacles and static obstacles on the road. In the process of generating one-way mapping information, only the traffic sign data representing traffic signs in the environmental data can be focused, and the obstacle data representing obstacles can be filtered out. Dynamic obstacles can include motor vehicles, pedestrians or non-motor vehicles. Static obstacles can include water horses, cones, etc. As joint map data, only dynamic obstacle data can be filtered out and static obstacle data retained. Therefore, the dynamic obstacle data is not universal and does not need to be retained when generating joint map data. Static obstacle data may last for a certain period of time and can be updated and stored as the vehicle travels on the corresponding road multiple times and as the static obstacles on the actual road change.
在本实施方式中，可以记录表示车辆所处环境中交通标识之间的位置关系。具体的，可以针对从环境数据具有交通标识的多个图像帧进行跟踪。例如，可以利用滤波的方式，对BEV视角下交通标识的类型、空间位置进行跟踪。对于跟踪的交通标识的一个图像帧，可以进行识别得到交通标识数据之后，再进一步的确认是否保留该图像帧。例如，由于障碍物遮挡之后，导致识别到的交通标识数据不够准确，可以放弃该图像帧的交通标识数据。通过放弃不够准确的交通标识数据，可以一定程度上减少出现错误的情况。具体的，滤波的算法可以包括但不限于：卡尔曼滤波、扩展卡尔曼滤波和粒子滤波等。In this embodiment, the positional relationship between traffic signs in the environment where the vehicle is located can be recorded. Specifically, tracking can be performed on multiple image frames with traffic signs from environmental data. For example, filtering can be used to track the type and spatial location of traffic signs from the BEV perspective. For an image frame of a tracked traffic sign, you can identify it and obtain the traffic sign data, and then further confirm whether to retain the image frame. For example, because the recognized traffic sign data is not accurate enough due to obstruction by obstacles, the traffic sign data of the image frame can be discarded. By discarding inaccurate traffic sign data, errors can be reduced to a certain extent. Specifically, filtering algorithms may include but are not limited to: Kalman filtering, extended Kalman filtering, particle filtering, etc.
在本实施方式中，可以针对局部的交通标识数据进行指定优化处理，以使得所述交通标识数据具有较为准确的处于三维空间的空间位置数据。具体的，可以基于BA局部优化算法，针对交通标识数据进行优化处理。进一步的，可以按照设定规则，针对多个交通标识数据添加相对位置标注。具体的，例如，针对表示停止线的交通标识数据与表示斑马线的交通标识数据，添加表示斑马线位于停止线前方的位置关系标注。例如，可以按照识别出的表示车道线的交通标识数据的先后顺序，进行顺序标注车道线位置关系。In this embodiment, specified optimization processing can be performed on the local traffic sign data, so that the traffic sign data has relatively accurate spatial position data in the three-dimensional space. Specifically, the traffic sign data can be optimized based on the BA local optimization algorithm. Furthermore, relative position annotations can be added to multiple traffic sign data according to the set rules. Specifically, for example, for the traffic sign data representing the stop line and the traffic sign data representing the zebra crossing, a positional relationship annotation indicating that the zebra crossing is located in front of the stop line is added. For example, lane line position relationships can be marked sequentially according to the sequence of recognized traffic sign data representing lane lines.
在本实施方式中，将从环境数据中识别的得到的交通标识信息，与按照采集时间对应的定位数据，整体形成单程建图信息。具体的，车辆的单程行驶会有确定的起点和终点。车辆到达终点之后，往往可能会驻车，并间隔一定时间之后，再开始下一趟的行程。如此，可以根据驾驶员对车辆的驾驶行为，将定位数据和环境数据划分为多个单程行程。如此，进而根据前文介绍生成单程行程的单程建图信息。In this implementation, the traffic sign information identified from the environmental data and the positioning data corresponding to the collection time are integrated to form one-way mapping information. Specifically, a vehicle's one-way journey will have a definite starting point and end point. After the vehicle reaches the end point, it may often park and wait for a certain period of time before starting the next trip. In this way, positioning data and environmental data can be divided into multiple one-way trips based on the driver's driving behavior of the vehicle. In this way, the one-way mapping information of the one-way trip is generated according to the previous introduction.
步骤S13：针对单程建图信息划分熟路组；其中，同一个熟路组中包括的单程建图信息中，单程建图信息的起点信息表示的位置之间符合第一设定距离条件，且终点信息表示的位置之间符合第二设定距离条件。Step S13: Divide familiar road groups for the one-way mapping information; among the one-way mapping information included in the same familiar road group, the positions represented by the starting point information of the one-way mapping information meet the first set distance condition, and the end point information The indicated positions meet the second set distance condition.
在本实施方式中，可以根据单程建图信息的起点信息和终点信息是否符合指定关联关系，进行划分熟路组。具体的，指定关联关系可以包括：单程建图信息的起点信息表示的起点之间符合第一设定距离关系、单程建图信息的终点信息表示的终点之间符合所述第二设定距离关系。当然，第一设定距离关系可以表示起点之间的距离需要满足的条件。第二设定距离关系可以表示终点之间的距离需要满足的条件。第一设定距离关系和第二设定距离关系可以相同，当然，也可以不同。具体的，第一设定距离关系和第二距离设定关系，可以依照实际需求进行设定。具体的，例如，第一设定距离关系为小于200米，第二设定距离关系也可以为小于200米。当然，第一设定距离关系可以为小于100米，第二设定距离关系可以为小于150米。在一些实施方式中，所述指定关联关系也可以包括路径重合度大于指定重合度阈值。具体的，多个路径信息的起点、途经点和终点总体形成路径之间的重合度。指定重合度阈值可以为70%，或75%、80%等等。在一些实施方式中，所述熟路组可以理解为车辆已经至少一次行驶于熟路组中单程建图信息对应的道路。当然，在一些实施方式中，也可以针对车辆行驶于单程建图信息对应道路的次数进行限定，在次数大于指定行驶次数阈值的情况下，才将单程建图信息划分入熟路组。In this embodiment, familiar road groups can be divided according to whether the starting point information and the ending point information of the one-way mapping information conform to the specified association relationship. Specifically, the specified association relationship may include: the starting points represented by the starting point information of the one-way mapping information comply with the first set distance relationship, and the end points represented by the end point information of the one-way mapping information comply with the second set distance relationship. . Of course, the first set distance relationship can represent the conditions that the distance between starting points needs to satisfy. The second set distance relationship can represent the conditions that the distance between the end points needs to meet. The first set distance relationship and the second set distance relationship may be the same, or of course, may be different. Specifically, the first set distance relationship and the second distance set relationship can be set according to actual needs. Specifically, for example, the first set distance relationship is less than 200 meters, and the second set distance relationship can also be less than 200 meters. Of course, the first set distance relationship may be less than 100 meters, and the second set distance relationship may be less than 150 meters. In some embodiments, the specified association relationship may also include that the path coincidence degree is greater than a specified coincidence degree threshold. Specifically, the starting points, passing points and end points of multiple path information collectively form the degree of overlap between paths. The specified coincidence threshold can be 70%, or 75%, 80%, etc. In some embodiments, the familiar road group can be understood as the vehicle has traveled on the road corresponding to the one-way mapping information in the familiar road group at least once. Of course, in some implementations, the number of times a vehicle travels on the road corresponding to the one-way mapping information can also be limited. Only when the number of times the vehicle travels is greater than a specified threshold of the number of trips, the one-way mapping information is divided into the familiar road group.
在一些实施方式中，不同熟路组可以用于表示一个驾驶员的使用场景。具体的，例如，对应驾驶员的上班场景的熟路组可以为上班熟路组，上班熟路组中包含的单程建图信息都表示驾驶员的上班路径。当然，也可以有下班熟路组，或者购物熟路组等等。In some embodiments, different familiar road groups may be used to represent a driver's usage scenario. Specifically, for example, the familiar road group corresponding to the driver's work scene may be the familiar road group to work, and the one-way mapping information contained in the familiar road group to work represents the driver's route to work. Of course, there can also be a group that is familiar with routes after get off work, or a group that is familiar with shopping routes, etc.
步骤S14：结合所述单程建图信息，建立并存储单程行程对应的语义地图，所述语义地图用于车辆在所述单程行程涉及的指定道路执行自动驾驶。Step S14: Combining the one-way mapping information, create and store a semantic map corresponding to the one-way trip. The semantic map is used for the vehicle to perform automatic driving on the designated road involved in the one-way trip.
在本实施方式中，指定道路可以是定位数据所对应的道路。具体的，定位数据所定位的位置属于所述指定道路。进一步的，定位数据连续变化所形成的行驶轨迹，经过的道路可以作为所述指定道路。具体的，例如，车辆从一个道路的起点驶入，从该道路的终点驶出，该道路可以作为所述指定道路。当然，车辆也可以从道路的中间位置从另一个道路驶入指定道路，此时建立的语义地图可以只涵盖指定道路被所述车辆途经的部分。In this embodiment, the designated road may be a road corresponding to the positioning data. Specifically, the location located by the positioning data belongs to the designated road. Furthermore, the driving trajectory formed by continuous changes in positioning data and the roads passed can be used as the designated roads. Specifically, for example, if a vehicle enters from the starting point of a road and exits from the end point of the road, the road can be used as the designated road. Of course, the vehicle can also enter the designated road from another road in the middle of the road. In this case, the semantic map established can only cover the part of the designated road passed by the vehicle.
语义地图可以是基于语义建图（semantic mapping）建立的，能够表示指定道路和交通标识信息的自建地图。建立针对指定道路的语义地图，如此，在车辆再次行驶在所述指定道路时，便可以基于语义地图执行自动驾驶。具体的，例如，在车辆行驶过程中，可以将语义地图作为一部分数据，输入至自动驾驶决策规划模块，如此，以进一步根据所述自动驾驶决策规划模块的运动规划控制所述车辆。在一些实施方式中，也可以仅针对熟路组进行语义建图，得出对应熟路组的语义地图。Semantic maps can be built based on semantic mapping and are self-built maps that can represent specified road and traffic sign information. Establish a semantic map for the designated road, so that when the vehicle drives on the designated road again, it can perform automatic driving based on the semantic map. Specifically, for example, during the driving process of the vehicle, the semantic map can be input as part of the data to the automatic driving decision planning module, so that the vehicle can be further controlled according to the motion planning of the automatic driving decision planning module. In some implementations, semantic mapping can also be performed only for familiar road groups to obtain a semantic map corresponding to the familiar road groups.
在一些实施方式中，可以生成表示定位数据对应的指定道路的道路对象；在道路对象中，添加表示定位数据对应的交通标识信息的交通标识对象。在本实施方式中，道路对象可以用于表示模拟所述指定道路。如此，车辆内部，便可以基于道路对象，控制车辆相对于指定道路的位置。道路对象可以包括表示指定道路的道路主体的主体对象，以及表示设置在道路主体上交通标识的交通标识对象。具体的，例如，主体对象可以按照一定比例模拟指定的道路的宽度和长度。交通标识对象可以表示的交通标识包括但不限于：地面交通标识和空中交通标识。地面交通标识可以包括车道边界标识、车道线等。空中交通标识可以包括红绿灯、限速牌等。具体的，例如，道路对象和交通标识对象等可以调用公开地图（OpenStreetMap，公开地图）建立。当然，也可以采用其他可以编辑地图软件建立道路对象和交通标识对象。In some implementations, a road object representing a designated road corresponding to the positioning data can be generated; in the road object, a traffic sign object representing the traffic sign information corresponding to the positioning data is added. In this embodiment, a road object may be used to represent the simulated designated road. In this way, inside the vehicle, the position of the vehicle relative to the specified road can be controlled based on the road object. The road object may include a subject object representing a road body of a specified road, and a traffic sign object representing a traffic sign set on the road body. Specifically, for example, the main object can simulate the width and length of the specified road according to a certain proportion. The traffic signs that the traffic sign object can represent include but are not limited to: ground traffic signs and air traffic signs. Ground traffic signs can include lane boundary signs, lane lines, etc. Air traffic signs can include traffic lights, speed limit signs, etc. Specifically, for example, road objects and traffic sign objects can be created by calling the public map (OpenStreetMap, public map). Of course, other map editing software can also be used to create road objects and traffic sign objects.
在本实施方式中，交通标识对象可以用于模拟交通标识信息表示的交通标识。交通标识信息对应有定位数据，如此交通标识对象也可以与定位数据相对应。如此，可以通过定位数据指示交通标识对象相对于道路对象的位置。如此，依照定位数据在道路对象上，添加交通标识对象。In this embodiment, the traffic sign object can be used to simulate the traffic sign represented by the traffic sign information. Traffic sign information corresponds to positioning data, so traffic sign objects can also correspond to positioning data. In this way, the position of the traffic sign object relative to the road object can be indicated by positioning data. In this way, a traffic sign object is added to the road object according to the positioning data.
在一些实施方式中，依照边界标识的定位数据可以在道路对象中添加表示边界标识的交通标识对象；将属于同一个熟路组中单程建图信息的指定道路之间，依照边界标识对齐相同的道路对象；为同一个熟路组中涉及的相同道路对象，调整为具有相同的交通标识对象。In some embodiments, according to the positioning data of the boundary identification, a traffic identification object representing the boundary identification can be added to the road object; between designated roads belonging to the one-way mapping information in the same familiar road group, the same roads can be aligned according to the boundary identification Object; it is the same road object involved in the same familiar road group, adjusted to have the same traffic sign object.
在本实施方式中，边界标识可以用于表示道路的边界。首先在道路对象中添加表示道路边界的交通标识信息。如此，便可以确定每个单程行程中，指定道路的边界标识对应的定位数据。然后，可以根据边界标识，将同一个熟路组中单程行程的道路对象对齐。此时，多个单程行程的同一个指定道路的道路对象之间，便可以对应有相同的位置数据。In this embodiment, the boundary mark may be used to represent the boundary of the road. First, add traffic sign information representing the road boundary to the road object. In this way, the positioning data corresponding to the boundary markers of the designated roads in each one-way trip can be determined. Road objects for one-way trips in the same familiar road group can then be aligned based on boundary markers. At this time, the road objects of the same designated road for multiple one-way trips can correspond to the same location data.
进一步的，便可以比较同一个熟路组中目标单程行程涉及的相同道路对象之间，所具有的交通标识对象是否存在差异。并在存在差异的情况下，可以调整一致。具体的，例如，可能会存在交通标识对象的缺少，或者，交通标识对象对应的定位数据不同的，在进行调整的过程中，可以相应增加交通标识对象，或者修改交通标识对象对应的定位数据。提升了熟路组中，道路对象设置交通标识对象的准确性。Furthermore, you can compare whether there are differences in traffic sign objects between the same road objects involved in the target one-way trip in the same familiar road group. And if there are differences, they can be adjusted to be consistent. Specifically, for example, there may be a lack of traffic sign objects, or the positioning data corresponding to the traffic sign objects is different. During the adjustment process, the traffic sign objects can be added accordingly, or the positioning data corresponding to the traffic sign objects can be modified. Improved the accuracy of setting traffic sign objects for road objects in the familiar road group.
在一些实施方式中，交通标识对象包括表示车道类型的车道类型标识对象。在所述语义地图中，车辆控制系统可以依照指定道路的车道类型标识对象，确定指定道路的车道数量；根据车道数量将指定道路划分子路段；其中，相邻子路段的车道数量不同；基于车道类型标识表示的车道类型，连接相邻子路段划分的车道。In some embodiments, the traffic identification object includes a lane type identification object representing a lane type. In the semantic map, the vehicle control system can determine the number of lanes of the designated road according to the lane type identification object of the designated road; divide the designated road into sub-sections according to the number of lanes; wherein the number of lanes in adjacent sub-sections is different; based on the lane The type of lane represented by the type identifier connects the lanes divided into adjacent sub-sections.
在本实施方式中，车道类型标识对象可以用于表示具有该车道类型标识对象的车道的车道类型。进一步的，可以依照车道类型标识对象的数量，确定指定道路中车道的数量。具体的，例如，存在两个分别表示左拐和直行的车道类型标识对象的情况下，可以确定指定道路具有两个车道。In this embodiment, the lane type identification object may be used to represent the lane type of the lane having the lane type identification object. Further, the number of lanes in the specified road can be determined according to the number of lane type identification objects. Specifically, for example, if there are two lane type identification objects representing left turns and straight ahead respectively, it can be determined that the designated road has two lanes.
在一些情况下，指定道路的车道数量可能会发生变化。具体的，例如，指定道路中可能具有并道，或者，增加车道的情况。此时，在车道数量发生变化的情况下，需要针对性的做好行为规划。如此，根据车道数量，将指定道路划分为多个子路段，使得车辆在该指定道路进行自动驾驶的过程中，可以依照指定道路的多个子路段做好的行为规划，减少在车道数量发生变化时导致车辆发生交通事故的概率。In some cases, the number of lanes on a designated roadway may change. Specifically, for example, the designated road may have a merge or add lanes. At this time, when the number of lanes changes, targeted behavior planning is required. In this way, the designated road is divided into multiple sub-sections according to the number of lanes, so that during the process of autonomous driving on the designated road, the vehicle can make behavioral plans according to the multiple sub-sections of the designated road, reducing the risk of accidents caused when the number of lanes changes. The probability of a vehicle being involved in a traffic accident.
如此，在语义地图中可以对应道路对象添加车道类型表示对象，以及为道路对象划分出多个车道。In this way, a lane type representation object can be added corresponding to the road object in the semantic map, and multiple lanes can be divided for the road object.
在一些实施方式中，在所述车辆行驶于所述指定道路的过程中，存储采集到的定位数据和环境数据；根据所述定位数据和所述环境数据更新所述指定道路的语义地图。如此，当车辆行驶于指定道路后，便可以根据最新采集的定位数据和环境数据，对语义地图进行更新。如此，可以提升语义地图的准确性。In some implementations, while the vehicle is driving on the designated road, the collected positioning data and environmental data are stored; and the semantic map of the designated road is updated according to the positioning data and the environmental data. In this way, when the vehicle drives on the designated road, the semantic map can be updated based on the latest collected positioning data and environmental data. In this way, the accuracy of the semantic map can be improved.
步骤S15：基于所述语义地图，生成所述单程建图信息对应的融合定位轨迹。Step S15: Based on the semantic map, generate a fused positioning trajectory corresponding to the one-way mapping information.
在本实施方式中，可以针对熟路组中的单程建图信息，生成对应的融合定位轨迹。具体的，可以获取根据环境数据得出的地面交通标识信息。具体的，可以获取表示车道线、道路边界线和停止线的地面交通标识信息。In this implementation, the corresponding fused positioning trajectory can be generated for the one-way mapping information in the familiar road group. Specifically, ground traffic sign information based on environmental data can be obtained. Specifically, ground traffic sign information representing lane lines, road boundary lines and stop lines can be obtained.
在本实施方式中，具体的，可以对应环境数据生成检测强度图。其中，在该检测强度图中，可以将对应车道线、道路边界线和停止线的像素分别设置为不同的指定像素值。可以理解为，将检测强度图中的多个像素划分了多个类别，同一个类别具有相同的像素值。不同类别的像素值不相同。具体的，例如，车道线、道路边界线和停止线分别划分为一类别，不属于前述类别的像素作为特殊类别。In this embodiment, specifically, a detection intensity map can be generated corresponding to the environmental data. Among them, in the detection intensity map, pixels corresponding to lane lines, road boundary lines and stop lines can be set to different specified pixel values respectively. It can be understood that multiple pixels in the detection intensity map are divided into multiple categories, and the same category has the same pixel value. The pixel values of different categories are different. Specifically, for example, lane lines, road boundary lines and stop lines are divided into one category respectively, and pixels that do not belong to the aforementioned categories are regarded as special categories.
可以使用卫星定位数据和惯导定位数据进行融合得到的定位数据，得到车辆在语义地图中的相对位姿T。并获取语义地图中，以相对位姿为中心，长宽分别为M米范围的道路对象。针对所述道路对象中，车道线，道路边界线和停止线分别作为一个类别，每个类别分别调整为指定像素值。得到道路对象的语义地图强度图。其中，语义地图强度图和所述检测强度图中，划分的类别相同，属于同一个类别的像素的像素值相同。同样，将不属于车道线，道路边界线和停止线的像素作为所述特殊类别。The positioning data obtained by fusing satellite positioning data and inertial navigation positioning data can be used to obtain the relative pose T of the vehicle in the semantic map. And obtain the road object in the semantic map, centered on the relative pose, with a length and width of M meters respectively. For the road object, lane lines, road boundary lines and stop lines are each regarded as a category, and each category is adjusted to a specified pixel value. Obtain the semantic map intensity map of road objects. Among them, the semantic map intensity map and the detection intensity map are divided into the same categories, and the pixel values of pixels belonging to the same category are the same. Likewise, pixels that do not belong to lane lines, road boundary lines, and stop lines are considered as said special categories.
在本实施方式中，可以使用位姿估计算法，基于检测强度图和语义地图强度图得到车辆的定位位姿。如此，基于多个环境数据的连续变化，将环境数据对应的相对位姿和定位位姿进行相互融合，得到车辆的融合行驶轨迹。其中位姿估计算法可以包括不限于迭代最近点算法（Iterative Closest Point，ICP），或者语义迭代最近点算法。In this embodiment, a pose estimation algorithm can be used to obtain the positioning pose of the vehicle based on the detection intensity map and the semantic map intensity map. In this way, based on the continuous changes of multiple environmental data, the relative pose and positioning pose corresponding to the environmental data are fused with each other to obtain the fused driving trajectory of the vehicle. The pose estimation algorithm may include but is not limited to iterative closest point algorithm (Iterative Closest Point, ICP) or semantic iterative closest point algorithm.
步骤S16：从所述融合行驶轨迹中解析得到所述车辆相对于道路的纵向和/或横向的目标速度分布，进而确定驾驶员的驾驶风格。Step S16: Analyze and obtain the longitudinal and/or transverse target speed distribution of the vehicle relative to the road from the fused driving trajectory, and then determine the driver's driving style.
在本实施方式中，可以根据所述融合行驶轨迹解析出车辆的横向速度数据和纵向速度数据分布，作为前述两个方向的目标速度分布数据。可以将目标速度分布数据与预先设定的驾驶风格进行匹配，得出对应驾驶员的目标驾驶风格。具体的，在解析目标速度分布数据的过程中，可以排除对应路口区域的融合行驶轨迹。通常路口区域的情况比较复杂，相对于其他道路段具有较强的特殊性，排除路口区域的融合行驶轨迹可以使得解析得到的目标速度分布数据可以更加准确的表示在非路口区域的道路段的速度情况，进而可以较为准确的确定目标驾驶风格。进一步的，每种驾驶风格都可以对应有基准速度分布数据，可以将目标速度分布数据分别与每种驾驶风格的基准速度分布数据进行比较，并将与速度分布数据最为相似的基准速度分布数据对应的驾驶风格，确定为目标驾驶风格。在一些实施方式中，可以根据横向速度数据和纵向速度数据计算方差，每种驾驶风格都可以对应有一个取值范围，不同驾驶风格之间取值范围不重叠，如此，可以将包含所述方差的取值范围对应的驾驶风格确定为目标驾驶风格。所述目标驾驶风格用于指导所述车辆在指定道路执行自动驾驶的过程中的速度分布；其中，不同驾驶风格的至少速度分布存在差异。In this embodiment, the lateral speed data and longitudinal speed data distribution of the vehicle can be analyzed based on the fused driving trajectory as the target speed distribution data in the two directions. The target speed distribution data can be matched with the preset driving style to obtain the corresponding driver's target driving style. Specifically, in the process of parsing the target speed distribution data, the fused driving trajectory corresponding to the intersection area can be excluded. Usually the situation in the intersection area is more complex and has strong particularity compared to other road segments. Excluding the fused driving trajectory of the intersection area can make the analyzed target speed distribution data more accurately represent the speed of the road segments in the non-intersection area. situation, and then the target driving style can be determined more accurately. Furthermore, each driving style can be corresponding to the benchmark speed distribution data. The target speed distribution data can be compared with the benchmark speed distribution data of each driving style respectively, and the benchmark speed distribution data most similar to the speed distribution data can be corresponding to the target speed distribution data. The driving style is determined as the target driving style. In some embodiments, the variance can be calculated based on the lateral speed data and the longitudinal speed data. Each driving style can have a corresponding value range. The value ranges between different driving styles do not overlap. In this way, the variance can be included The driving style corresponding to the value range is determined as the target driving style. The target driving style is used to guide the speed distribution of the vehicle during automatic driving on a designated road; where at least the speed distribution of different driving styles is different.
进一步的，在一些实施方式中，可以根据目标速度分布数据确定导航路径信息涉及道路的驾驶难度。所述驾驶难度可以包括高难度、普通难度和低难度。具体的，例如，如果一个道路的纵向的目标速度分布数据小于一定的速度阈值，可以判定该道路为经常拥堵路段，可以认定该路为高难度。如果一个道路的纵向目标速度分布数据较为均匀的维持在一个较快速度，可以认定该道路为低难度。在一些实施方式中，可以针对路口区域单独识别驾驶难度。具体的，例如，根据目标融合行驶轨迹得出路口区域横向的目标速度分布数据经常变化，说明车辆在经过该路口区域时面临较复杂的驾驶情况，判定该路口区域的驾驶难度为高难度。所属领域技术人员，可以根据实际需要进行设定难度规则，以指定导航路径信息涉及道路的驾驶难度。Further, in some embodiments, the driving difficulty of the road involved in the navigation path information may be determined based on the target speed distribution data. The driving difficulty may include high difficulty, normal difficulty and low difficulty. Specifically, for example, if the longitudinal target speed distribution data of a road is less than a certain speed threshold, the road can be determined to be a frequently congested road section, and the road can be determined to be highly difficult. If the longitudinal target speed distribution data of a road is relatively uniform and maintains a relatively fast speed, the road can be deemed to be of low difficulty. In some implementations, driving difficulty may be identified separately for intersection areas. Specifically, for example, based on the target fusion driving trajectory, the lateral target speed distribution data of the intersection area often changes, indicating that the vehicle faces a more complex driving situation when passing through the intersection area, and the driving difficulty of the intersection area is determined to be high. Those skilled in the art can set difficulty rules according to actual needs to specify the driving difficulty of the roads involved in the navigation path information.
请参阅图2，所述影子模式验证阶段，可以基于所述融合行驶轨迹调整所述车辆的自动驾驶决策规划模块。具体的，车辆控制系统可以从所述融合行驶轨迹中解析得到所述车辆相对于道路的纵向和/或横向的目标速度分布；调整所述自动驾驶决策规划模块，以使所述车辆自动驾驶于所述熟路组对应的道路的过程中，所述车辆的速度分布趋于所述目标速度分布。具体的，所述影子模式验证阶段可以包括以下步骤。Please refer to Figure 2. In the shadow mode verification stage, the automatic driving decision planning module of the vehicle can be adjusted based on the fused driving trajectory. Specifically, the vehicle control system can analyze the longitudinal and/or transverse target speed distribution of the vehicle relative to the road from the fused driving trajectory; adjust the automatic driving decision planning module so that the vehicle automatically drives in During the course of the roads corresponding to the familiar road group, the speed distribution of the vehicle tends to the target speed distribution. Specifically, the shadow mode verification phase may include the following steps.
步骤S21：在车辆行驶过程中，结合车辆的卫星定位数据和惯导定位数据，得到车辆的定位数据。Step S21: During the driving process of the vehicle, the vehicle's positioning data is obtained by combining the vehicle's satellite positioning data and inertial navigation positioning data.
在本实施方式中，车辆控制系统可以在车辆行驶过程中，不断结合车辆的卫星定位数据和惯导定位数据，以得到由补全的卫星定位数据形成的定位数据，具体的，可以参照前述实施方式介绍，不再赘述。In this embodiment, the vehicle control system can continuously combine the vehicle's satellite positioning data and inertial navigation positioning data during the driving process of the vehicle to obtain positioning data formed by the complementary satellite positioning data. For details, please refer to the aforementioned implementation. The method is introduced without going into details.
步骤S22：基于采集到的环境数据，进行障碍物和交通标识识别，得到表示障碍物的障碍物数据，和表示交通标识的交通标识数据。Step S22: Based on the collected environmental data, identify obstacles and traffic signs to obtain obstacle data representing obstacles and traffic sign data representing traffic signs.
在本实施方式中，具体的，可以基于环境数据进行障碍物识别、空中交通标识识别、地面交通标识识别等。具体的，可以参见前述实施方式对照解释，不再赘述。In this embodiment, specifically, obstacle recognition, air traffic sign recognition, ground traffic sign recognition, etc. can be performed based on environmental data. For details, please refer to the foregoing embodiments for comparative explanation, and no further description will be given.
步骤S23：根据所述车辆的定位数据，在存储的语义地图中读取指定位置范围内的局部语义地图。Step S23: According to the positioning data of the vehicle, read the local semantic map within the specified location range from the stored semantic map.
在本实施方式中，车辆控制系统可以具有语义地图引擎，所述语义地图引擎可以根据定位数据从存储的语义地图中读取所述车辆的定位数据附近的局部语义地图，从而可以一定程度上减少数据处理量。In this embodiment, the vehicle control system may have a semantic map engine. The semantic map engine may read the local semantic map near the vehicle's positioning data from the stored semantic map according to the positioning data, thereby reducing the number of problems to a certain extent. Data processing volume.
在本实施方式中，局部语义地图可以包括指定位置范围内，构成语义地图的地图语义要素，和语义要素的位置。具体的，例如，地图语义要素可以包括道路对象、交通标识对象等，以及位置关系。所述指定位置范围可以为指定的距离范围。具体的，例如，指定位置范围为相对于所述定位数据的距离为300米以内。当然，并不限于300米，还可以为400米或500米。In this embodiment, the local semantic map may include map semantic elements constituting the semantic map within a specified location range, and the locations of the semantic elements. Specifically, for example, map semantic elements may include road objects, traffic sign objects, etc., as well as location relationships. The specified location range may be a specified distance range. Specifically, for example, the specified location range is within 300 meters relative to the positioning data. Of course, it is not limited to 300 meters, and can also be 400 meters or 500 meters.
步骤S24：结合所述局部语义地图、环境数据和定位数据，生成车辆的融合行驶轨迹。Step S24: Combine the local semantic map, environmental data and positioning data to generate a fused driving trajectory of the vehicle.
在本实施方式中，可以将车辆行驶过程中，采集到环境数据和定位数据，与局部语义地图进行融合处理，生成车辆的融合行驶轨迹。具体的，可以参照前述实施方式对照解释，不再赘述。In this embodiment, the environmental data and positioning data collected during the driving of the vehicle can be fused with the local semantic map to generate the fused driving trajectory of the vehicle. Specifically, reference may be made to the foregoing embodiments for comparative explanation, and no further description will be given.
步骤S25：基于从环境数据中识别出的地面交通标识数据，进行局部语义建图，得到局部地图。Step S25: Based on the ground traffic sign data identified from the environmental data, perform local semantic mapping to obtain a local map.
在本实施方式中，所述局部地图可以用于表示车辆所处环境中地面交通标识之间的位置关系。具体的，可以针对从环境数据具有地面交通标识的多个图像帧进行跟踪。例如，可以利用滤波的方式，对BEV视角下地面交通标识的类型、空间位置进行跟踪。对于跟踪的地面交通标识可以进行识别得到交通标识数据之后，再进一步的确认是否保留。例如，针对一些由于障碍物遮挡导致交通标识数据，相较于之前识别到的交通标识数据来说，可能是同一个交通标识，但由于障碍物遮挡之后，导致识别到的交通标识数据不够准确，可以放弃该图像帧的交通标识数据。通过放弃不够准确的交通标识数据，可以一定程度上减少出现错误的情况。具体的，滤波的算法可以包括但不限于：卡尔曼滤波、扩展卡尔曼滤波和粒子滤波等。In this embodiment, the local map can be used to represent the positional relationship between ground traffic signs in the environment where the vehicle is located. Specifically, tracking can be performed on multiple image frames with ground traffic identification from environmental data. For example, filtering can be used to track the type and spatial location of ground traffic signs from the BEV perspective. The tracked ground traffic signs can be identified and the traffic sign data can then be further confirmed to be retained. For example, for some traffic sign data due to obstruction by obstacles, compared with the previously recognized traffic sign data, it may be the same traffic sign. However, due to obstruction by obstacles, the recognized traffic sign data is not accurate enough. The traffic sign data of the image frame can be discarded. By discarding inaccurate traffic sign data, errors can be reduced to a certain extent. Specifically, filtering algorithms may include but are not limited to: Kalman filtering, extended Kalman filtering, particle filtering, etc.
在本实施方式中，可以针对局部地图中的交通标识数据进行指定优化处理，以使得所述交通标识数据具有较为准确的处于三维空间的空间位置数据。具体的，可以基于BA局部优化算法，针对交通标识数据进行优化处理。进一步的，可以按照设定规则，针对多个交通标识数据添加相对位置标注。具体的，例如，针对表示停止线的交通标识数据与表示斑马线的交通标识数据，添加表示斑马线位于停止线前方的位置关系标注。例如，可以按照识别出的表示车道线的交通标识数据的先后顺序，进行顺序标注车道线位置关系。In this embodiment, designated optimization processing can be performed on the traffic sign data in the local map, so that the traffic sign data has relatively accurate spatial position data in the three-dimensional space. Specifically, the traffic sign data can be optimized based on the BA local optimization algorithm. Furthermore, relative position annotations can be added to multiple traffic sign data according to the set rules. Specifically, for example, for the traffic sign data representing the stop line and the traffic sign data representing the zebra crossing, a positional relationship annotation indicating that the zebra crossing is located in front of the stop line is added. For example, lane line position relationships can be marked sequentially according to the sequence of recognized traffic sign data representing lane lines.
步骤S26：根据表示障碍物的障碍物数据和所述局部地图进行轨迹预测，得到轨迹预测结果。Step S26: Perform trajectory prediction based on the obstacle data representing obstacles and the local map to obtain a trajectory prediction result.
在本实施方式中，可以将障碍物数据和局部地图输入给轨迹预测模块，以得到轨迹预测模块输出的轨迹预测结果。具体的，例如，轨迹预测模块可以采用TNT或者DenseTNT实现。具体的，例如，轨迹预测模块可以基于局部语义地图获取表示车辆的所处车道的车道数据、车道数量、距离下一路口的距离、下一路口的导航指令信息、路口车道分配关系、动态障碍物数据和静态障碍物数据等，得出车辆的多个预测轨迹。导航指令信息可以包括但不限于直行、左转、右转和掉头等。路口车道分配关系是指车道被允许的行驶行为。例如，最右侧车道为右转车道，中间为直行车道，最左侧车道为左转车道。轨迹预测模块可以根据前述输入的多个参数进行多轨迹预测，并进一步的为多个轨迹进行打分，并选择得分最高的预测轨迹作为轨迹预测结果。In this embodiment, the obstacle data and the local map can be input to the trajectory prediction module to obtain the trajectory prediction result output by the trajectory prediction module. Specifically, for example, the trajectory prediction module can be implemented using TNT or DenseTNT. Specifically, for example, the trajectory prediction module can obtain lane data representing the lane where the vehicle is located, the number of lanes, the distance to the next intersection, the navigation instruction information of the next intersection, intersection lane allocation relationships, and dynamic obstacles based on the local semantic map. Data and static obstacle data, etc., to obtain multiple predicted trajectories of the vehicle. Navigation instruction information may include, but is not limited to, go straight, turn left, turn right, and make a U-turn. The lane allocation relationship at an intersection refers to the permitted driving behavior of the lane. For example, the rightmost lane is the right turn lane, the middle is the through lane, and the leftmost lane is the left turn lane. The trajectory prediction module can perform multi-trajectory prediction based on the multiple input parameters mentioned above, and further score multiple trajectories, and select the predicted trajectory with the highest score as the trajectory prediction result.
步骤S27：基于所述融合行驶轨迹数据、所述局部语义地图的道路对象、所述车辆的导航路径信息和所述轨迹预测结果，进行所述车辆的虚拟行为规划和虚拟运动规划。Step S27: Carry out virtual behavior planning and virtual motion planning of the vehicle based on the fused driving trajectory data, the road objects of the local semantic map, the navigation path information of the vehicle, and the trajectory prediction results.
在一些实施方式中，车辆控制系统可以将车辆的融合行驶轨迹数据、局部语义地图的道路对象、车辆的导航路径信息和轨迹预测结果，输入至车辆控制系统的自动驾驶决策规划模块，由所述自动驾驶决策规划模块输出虚拟行为规划和虚拟运动规划。自动驾驶决策规划模块输出的虚拟行为规划和虚拟运动规划可以用于实现所述轨迹预测结果。具体的，例如，若车辆所处的车道与轨迹预测结果表示的目标车道不符，自动驾驶决策规划模块需要给出需要变道的虚拟行为规划。如果，存在表示限速的交通标识信息，自动驾驶决策规划模块可以根据车速判断是否输出需要降速或提速的虚拟行为规划。自动驾驶决策规划模块可以根据红绿灯的状态和车辆到达路口的距离，输出决定降速或刹车的虚拟行为规划。对于过路口，自动驾驶决策规划模块可以基于设定规则，输出转入下一道路之后的目标车道的虚拟行为规划。例如，设定规则可以包括：默认右转之后进入右侧第一个车道，左转到右侧最后一个车道，直行进入与车道的横向距离较近的车道等。进一步的，还可以结合驾驶员的驾驶风格、具体的，在进行决策判断时，在多个规划中选择与车主驾驶风格相近似的规划。针对一些，被标记为驾驶难度大的路段，自动驾驶决策规划模块可以输出偏谨慎的虚拟行为规划。具体的，偏谨慎的虚拟行为规划可以为被认定具有较高安全性的虚拟行为规划。例如，在车辆前方具有动态障碍物的情况下，自动驾驶决策规划模块可以输出控制车辆停车等待的虚拟行为规划。在车辆行驶于路口的情况下，自动驾驶决策规划模块可以输出控制车辆避免变道的虚拟行为规划。In some embodiments, the vehicle control system can input the vehicle's fused driving trajectory data, the road objects of the local semantic map, the vehicle's navigation path information and trajectory prediction results into the automatic driving decision planning module of the vehicle control system. The autonomous driving decision planning module outputs virtual behavior planning and virtual motion planning. The virtual behavior planning and virtual motion planning output by the autonomous driving decision planning module can be used to achieve the trajectory prediction results. Specifically, for example, if the lane the vehicle is in does not match the target lane represented by the trajectory prediction result, the autonomous driving decision planning module needs to provide a virtual behavior plan that requires lane change. If there is traffic sign information indicating a speed limit, the autonomous driving decision planning module can determine whether to output a virtual behavior plan that requires slowing down or speeding up based on the vehicle speed. The autonomous driving decision planning module can output a virtual behavior plan that decides to slow down or brake based on the status of the traffic light and the distance of the vehicle to the intersection. For intersections, the autonomous driving decision planning module can output a virtual behavior plan for the target lane after turning into the next road based on set rules. For example, setting rules may include: entering the first lane on the right after turning right by default, turning left to the last lane on the right, going straight into a lane that is closer to the lane laterally, etc. Furthermore, the driver's driving style and specific driving style can also be combined to select a plan that is similar to the car owner's driving style among multiple plans when making decision-making judgments. For some road sections marked as difficult to drive, the autonomous driving decision planning module can output a cautious virtual behavior plan. Specifically, the more cautious virtual behavior planning can be the virtual behavior planning that is considered to have higher security. For example, when there are dynamic obstacles in front of the vehicle, the autonomous driving decision planning module can output a virtual behavior plan that controls the vehicle to stop and wait. When the vehicle is driving at an intersection, the autonomous driving decision planning module can output a virtual behavior plan that controls the vehicle to avoid changing lanes.
进一步的，可以自动驾驶决策规划模块可以基于虚拟行为规划的规划结果，进行虚拟运动规划。具体的，例如，可以结合虚拟行为规划、障碍物数据进行虚拟运动规划。具体的，例如，自动驾驶决策规划模块可以采取横纵分离的虚拟运动规划，或者横纵融合的虚拟运动规划。在一些实施方式中，自动驾驶决策规划模块可以采用MPC（Model PredictiveControl，模型预测控制）进行虚拟运动规划。Furthermore, the autonomous driving decision planning module can perform virtual motion planning based on the planning results of virtual behavior planning. Specifically, for example, virtual motion planning can be performed in combination with virtual behavior planning and obstacle data. Specifically, for example, the autonomous driving decision planning module can adopt virtual motion planning that separates horizontal and vertical directions, or virtual motion planning that integrates horizontal and vertical directions. In some implementations, the autonomous driving decision planning module can use MPC (Model Predictive Control, Model Predictive Control) for virtual motion planning.
在本实施方式中，自动驾驶决策规划模块可以每间隔一个时间周期，输出虚拟行为规划和虚拟运动规划。时间周期可以为80毫秒、100毫秒、150毫秒等。In this implementation, the autonomous driving decision-making and planning module can output virtual behavior planning and virtual motion planning at intervals of one time period. The time period can be 80 milliseconds, 100 milliseconds, 150 milliseconds, etc.
步骤S28：将基于所述虚拟行为规划和所述虚拟运动规划的虚拟控制数据，与驾驶员的实际控制数据进行比较，得出差异数据，以根据所述差异数据修改所述自动驾驶决策规划模块。Step S28: Compare the virtual control data based on the virtual behavior planning and the virtual motion planning with the driver's actual control data to obtain difference data, so as to modify the automatic driving decision planning module according to the difference data. .
在本实施方式中，差异数据可以用于表示基于虚拟控制数据控制车辆，与驾驶员控制车辆之间的差异。具体的，差异数据可以包括行为规划差异数据和运动规划差异数据。In this embodiment, the difference data may be used to represent the difference between controlling the vehicle based on virtual control data and controlling the vehicle based on the driver. Specifically, the difference data may include behavior planning difference data and movement planning difference data.
在本实施方式中，车辆控制系统可以基于虚拟行为规划和虚拟运动规划执行针对车辆的虚拟控制，得出虚拟控制数据。具体的，例如，车辆控制系统可以基于虚拟行为规划，模拟车辆的横向控制和纵向控制得到虚拟横向控制数据和虚拟纵向控制数据。在本实施方式中，车辆控制系统可以结合车辆运动学，针对车辆的虚拟横向控制数据和虚拟纵向控制数据表示的控制进行模拟。进一步的，可以根据虚拟控制数据模拟针对车辆的控制，依照前文实施方式介绍的技术方案形成车辆的目标融合行驶轨迹数据的方式，生成虚拟行驶轨迹数据。具体的，可以针对每个所述时间周期的虚拟控制数据，得出该虚拟控制数据对应的虚拟行驶轨迹数据。In this embodiment, the vehicle control system can perform virtual control for the vehicle based on virtual behavior planning and virtual motion planning to obtain virtual control data. Specifically, for example, the vehicle control system can simulate the lateral control and longitudinal control of the vehicle based on virtual behavior planning to obtain virtual lateral control data and virtual longitudinal control data. In this embodiment, the vehicle control system can simulate the control represented by the virtual lateral control data and the virtual longitudinal control data of the vehicle in combination with the vehicle kinematics. Further, the control of the vehicle can be simulated based on the virtual control data, and the vehicle's target fusion driving trajectory data can be formed according to the technical solution introduced in the previous embodiment to generate virtual driving trajectory data. Specifically, for the virtual control data of each time period, the virtual driving trajectory data corresponding to the virtual control data can be obtained.
在本实施方式中，车辆控制系统可以读取驾驶员实际控制所述车辆产生的实际横向控制数据和实际纵向控制数据，进而生成差异数据。具体的如此，可以针对虚拟横向控制数据和实际横向控制数据进行比较，针对虚拟纵向控制数据和实际纵向控制数据进行比较，根据比较结果得出所述虚拟行为规划的虚拟横向差异数据和虚拟纵向差异数据。进一步的，分别根据所述虚拟横向差异数据和所述虚拟纵向差异数据表示的差异，调整所述自动驾驶决策规划模块。具体的，例如，可以将虚拟横向控制数据表示的控制方向，与实际横向控制数据针对车辆的控制方向比较，在二者方向相反的情况下，记录表示方向差异的差异数据。例如，可以将虚拟纵向控制数据表示的加速或减速，与实际纵向控制数据针对车辆的加速或减速控制进行比较，在二者不一致的情况下，记录表示加减速差异的差异数据。当然，在一些实施方式中，差异数据可以表示存在方向差异的差异次数。In this embodiment, the vehicle control system can read the actual lateral control data and actual longitudinal control data generated by the driver's actual control of the vehicle, and then generate difference data. Specifically, the virtual lateral control data and the actual lateral control data can be compared, the virtual longitudinal control data and the actual longitudinal control data can be compared, and the virtual lateral difference data and the virtual vertical difference of the virtual behavior planning can be obtained according to the comparison results. data. Further, the automatic driving decision planning module is adjusted according to the difference represented by the virtual lateral difference data and the virtual longitudinal difference data respectively. Specifically, for example, the control direction represented by the virtual lateral control data can be compared with the control direction of the vehicle by the actual lateral control data. If the two directions are opposite, the difference data indicating the difference in direction is recorded. For example, the acceleration or deceleration represented by the virtual longitudinal control data can be compared with the acceleration or deceleration control of the vehicle by the actual longitudinal control data. If the two are inconsistent, difference data indicating the difference in acceleration and deceleration is recorded. Of course, in some embodiments, the difference data may represent the number of differences for which directional differences exist.
在一些实施方式中，可以根据差异数据调整自动驾驶决策规划模块，以使得虚拟控制数据可以达成与驾驶员的实际控制数据之间，减少虚拟行为规划与驾驶员的实际驾驶行为之间的偏差。具体的，虚拟控制数据可以形成车辆的虚拟控制行驶轨迹数据，所述虚拟控制行驶轨迹数据可以具有横向速度分布和纵向速度分布，可以通过调整自动驾驶决策规划模块使得虚拟控制行驶轨迹数据的横向速度分布和纵向速度分布，与驾驶员的实际驾驶车辆形成的横向速度分布和纵向速度分布趋于相同。In some implementations, the automatic driving decision planning module can be adjusted according to the difference data, so that the virtual control data can be compared with the driver's actual control data, and the deviation between the virtual behavior planning and the driver's actual driving behavior can be reduced. Specifically, the virtual control data can form the virtual control driving trajectory data of the vehicle. The virtual control driving trajectory data can have lateral speed distribution and longitudinal speed distribution. The lateral speed of the virtual control driving trajectory data can be controlled by adjusting the automatic driving decision planning module. The distribution and longitudinal velocity distribution tend to be the same as the lateral velocity distribution and longitudinal velocity distribution formed by the driver's actual driving vehicle.
在本实施方式中，在一个时间周期结束之后，如果认定该时间周期的虚拟行为规划与驾驶员的实际驾驶行为之间存在差异，可以将该时间周期作为目标时间周期，可以记录将该目标时间周期对应的目标融合行驶轨迹数据和轨迹预测结果。如此，自动驾驶决策规划模块可以将当前输入的目标融合行驶轨迹数据和轨迹预测结果，与之前记录的目标时间周期对应的目标融合行驶轨迹数据和轨迹预测结果，进行比较。在当前的目标融合行驶轨迹数据和轨迹预测结果，与记录的目标融合行驶轨迹数据和轨迹预测结果较为近似的情况下，目标时间周期对应的差异数据作为所述自动决策规划模块的一个输入量，以使得自动决策规划模块在生成虚拟行为规划时，可以结合之前的差异数据，如此使得到的虚拟行为规划可以更加接近驾驶员的实际驾驶行为。在一些实施方式中，因为当前的目标融合行驶轨迹数据和轨迹预测结果仅仅可能是与目标时间周期的目标融合行驶轨迹数据和轨迹预测结果相近似，但并不相同。为了避免自动驾驶决策规划模块调整幅度过大，反而导致输出的虚拟行为规划与驾驶员的实际驾驶行为更大的差异。可以将目标时间周期的差异数据进行加权之后，输入给自动驾驶决策规划模块。具体的，例如，针对差异数据进行加权的权值可以为0.2、0.3或0.35等。In this embodiment, after the end of a time period, if it is determined that there is a difference between the virtual behavior plan of the time period and the driver's actual driving behavior, the time period can be used as the target time period, and the target time can be recorded. The target corresponding to the period fuses the driving trajectory data and trajectory prediction results. In this way, the autonomous driving decision planning module can compare the currently input target fusion driving trajectory data and trajectory prediction results with the target fusion driving trajectory data and trajectory prediction results corresponding to the previously recorded target time period. When the current target fusion driving trajectory data and trajectory prediction results are relatively similar to the recorded target fusion driving trajectory data and trajectory prediction results, the difference data corresponding to the target time period is used as an input quantity of the automatic decision planning module, This allows the automatic decision-making planning module to combine the previous differential data when generating the virtual behavior plan, so that the resulting virtual behavior plan can be closer to the driver's actual driving behavior. In some embodiments, the current target fused driving trajectory data and trajectory prediction results may only be similar to, but not identical to, the target fused driving trajectory data and trajectory prediction results of the target time period. In order to avoid excessive adjustment of the automatic driving decision planning module, which will lead to a greater difference between the output virtual behavior planning and the driver's actual driving behavior. The difference data of the target time period can be weighted and then input to the autonomous driving decision planning module. Specifically, for example, the weight for weighting differential data may be 0.2, 0.3, or 0.35, etc.
在本实施方式中，如果认定车辆行驶于该导航路径信息表示道路时，自动驾驶决策规划模块输出的虚拟行为规划与驾驶员的实际驾驶行为之间没有差异。可以进一步判断是否需要进行虚拟运动规划的修正。具体的，可以根据每个时间周期的虚拟行为规划，模拟控制车辆得到车辆在每个时间周期的虚拟控制行驶轨迹数据，以及获得每个时间周期对应的车辆实际的目标融合行驶轨迹数据。如此，进一步计算虚拟控制行驶轨迹数据与车辆实际的目标融合行驶轨迹数据之间的绝对位姿误差（absolute pose error，APE），如此，可以根据绝对位姿误差针对自动驾驶决策规划模块进行修订。当然，在一些实施方式中，可以在绝对位姿误差的取值大于一个设定位姿误差阈值的情况下，才针对自动驾驶决策规划模块进行修订。当然，在一些实施方式中，还可以计算虚拟控制行驶轨迹数据和车辆实际的目标融合行驶轨迹数据之间的绝对轨迹误差（absolute trajectory error，ATE）、相对位姿误差（relative pose error，RPE）和相对轨迹误差（relative trajectory error，RTE），并相应的修正自动驾驶决策规划模块。In this embodiment, if it is determined that the vehicle is traveling on the road represented by the navigation path information, there is no difference between the virtual behavior plan output by the autonomous driving decision planning module and the driver's actual driving behavior. It can be further determined whether the virtual motion planning needs to be corrected. Specifically, according to the virtual behavior planning of each time period, the vehicle can be simulated and controlled to obtain the virtual control driving trajectory data of the vehicle in each time period, and the actual target fusion driving trajectory data of the vehicle corresponding to each time period can be obtained. In this way, the absolute pose error (APE) between the virtual control driving trajectory data and the vehicle's actual target fusion driving trajectory data is further calculated. In this way, the automatic driving decision planning module can be revised based on the absolute pose error. Of course, in some implementations, the automatic driving decision planning module can be revised only when the value of the absolute pose error is greater than a set pose error threshold. Of course, in some implementations, the absolute trajectory error (ATE) and relative pose error (RPE) between the virtual control driving trajectory data and the vehicle's actual target fusion driving trajectory data can also be calculated. and relative trajectory error (RTE), and correct the autonomous driving decision planning module accordingly.
步骤S29：根据所述差异数据，生成单程建图信息对应的自动驾驶信心信息。Step S29: Generate autonomous driving confidence information corresponding to the one-way mapping information based on the difference data.
在本实施方式中，可以根据所述差异数据针对被划分为熟路组的单程建图信息，生成对应单程建图信息的自动驾驶信心信息。具体的，例如，可以根据车辆重复行驶在同一个单程建图信息表示的道路的差异数据，按照信心指数生成规则生成单程建图信息的自动驾驶信心信息。具体的，例如，信心指数生成规则可以包括：车辆行驶于所述单程建图信息表示道路的差异数据等于第一指定阈值，认定车辆行驶于所述单程建图信息表示道路的自动驾驶信心信息为优秀；车辆行驶于所述单程建图信息表示道路的差异数据大于第一指定阈值，且小于等于第二指定阈值的情况下，认定车辆行驶于所述单程建图信息表示道路的自动驾驶信心信息为良好；车辆行驶于所述单程建图信息的差异数据大于第二指定阈值，且小于等于第三阈值的情况下，认定车辆行驶于所述单程建图信息表示道路的自动驾驶信心信息为正常；车辆行驶于所述单程建图信息的差异数据大于第三阈值的情况下，认定车辆行驶于所述单程建图信息表示道路的自动驾驶信心信息为差。在一些实施方式中，可以根据差异数据，统计车辆行驶于单程建图信息表示道路过程中，每公里出现虚拟行为规划与驾驶员的实际驾驶行为之间的差异次数。如此，第一指定阈值可以为0，在差异次数等于第一指定阈值的情况下，可以表示该单程建图信息对应的自动驾驶信心信息为优秀，此时表示自动驾驶能力为优秀。第二指定阈值可以为1，在差异次数大于0，并小于等于1时，可以表示该单程建图信息对应的自动驾驶信心信息为良好，此时表示自动驾驶能力为良好。第三指定阈值可以为2，在差异次数大于1，并小于等于2时，可以表示该单程建图信息对应的自动驾驶信心信息为正常，此时表示自动驾驶能力为正常。当差异次数大于2时，可以表示该单程建图信息对应的自动驾驶信心信息为差，此时表示自动驾驶能力为差。In this embodiment, automatic driving confidence information corresponding to the one-way mapping information can be generated based on the difference data for the one-way mapping information divided into familiar road groups. Specifically, for example, the autonomous driving confidence information of the one-way mapping information can be generated according to the confidence index generation rules based on the difference data of the vehicle repeatedly driving on the same road represented by the one-way mapping information. Specifically, for example, the confidence index generation rule may include: the difference data of the vehicle traveling on the road represented by the one-way mapping information is equal to the first specified threshold, and it is determined that the self-driving confidence information of the vehicle traveling on the road represented by the one-way mapping information is Excellent; when the difference data of the vehicle driving on the road represented by the one-way mapping information is greater than the first specified threshold and less than or equal to the second designated threshold, it is determined that the vehicle is driving on the road represented by the one-way mapping information and the self-driving confidence information is good; when the difference data of the vehicle traveling on the one-way mapping information is greater than the second specified threshold and less than or equal to the third threshold, it is determined that the self-driving confidence information of the vehicle traveling on the road represented by the one-way mapping information is normal. ; When the difference data of the vehicle traveling on the one-way mapping information is greater than the third threshold, it is determined that the self-driving confidence information of the vehicle traveling on the road represented by the one-way mapping information is poor. In some implementations, the difference data can be used to count the number of differences between the virtual behavior plan and the driver's actual driving behavior per kilometer during the vehicle's driving on the road represented by the one-way mapping information. In this way, the first specified threshold can be 0. When the number of differences is equal to the first specified threshold, it can mean that the autonomous driving confidence information corresponding to the one-way mapping information is excellent, which means that the autonomous driving capability is excellent. The second specified threshold can be 1. When the number of differences is greater than 0 and less than or equal to 1, it can mean that the self-driving confidence information corresponding to the one-way mapping information is good, which means that the self-driving capability is good. The third specified threshold can be 2. When the number of differences is greater than 1 and less than or equal to 2, it can mean that the self-driving confidence information corresponding to the one-way mapping information is normal, which means that the self-driving capability is normal. When the number of differences is greater than 2, it can mean that the autonomous driving confidence information corresponding to the one-way mapping information is poor, which means that the autonomous driving capability is poor.
在一些实施方式中，自动驾驶信心信息可以为具体的数值。具体的，可以设定自动驾驶信心信息的最大值为100。当差异次数为0时，可以认定单程建图信息对应的自动驾驶信心信息为100，此时表示自动驾驶能力为优秀。当差异次数大于0并小于等于1时，可以映射至数值小于100并大于等于90的取值区间，此时表示自动驾驶能力为良好。当差异次数大于1并小于等于2时，可以映射至数值小于90并大于等于80，此时表示自动驾驶能力为正常。当差异次数大于2时，可以映射至数值小于80，此时表示自动驾驶能力为差。在一些实施方式中，可以结合驾驶难度，为自动驾驶信心信息的数值设置权值。具体的，单程建图信息的自动驾驶难度为高难度时，可以为前述计算得出的自动驾驶信心信息设置第一权值。当单程建图信息的自动驾驶难度为普通难度时，可以为自动驾驶信心信息设置第二权值。当单程建图信息的自动驾驶难度为低难度时，可以为自动驾驶信心信息设置第三权值。其中，第一权值大于第二权值大于第三权值。在自动驾驶信心信息为100时，可以无需设置权值。如此，可以将自动驾驶的难度作用至自动驾驶信心信息，以使得可以更加安全、准确的表征车辆执行自动驾驶的安全性。In some implementations, the autonomous driving confidence information may be a specific numerical value. Specifically, the maximum value of autonomous driving confidence information can be set to 100. When the number of differences is 0, it can be determined that the autonomous driving confidence information corresponding to the one-way mapping information is 100, which indicates that the autonomous driving capability is excellent. When the number of differences is greater than 0 and less than or equal to 1, it can be mapped to a value range of less than 100 and greater than or equal to 90, which indicates that the autonomous driving capability is good. When the number of differences is greater than 1 and less than or equal to 2, it can be mapped to a value less than 90 and greater than or equal to 80, which indicates that the autonomous driving capability is normal. When the number of differences is greater than 2, it can be mapped to a value less than 80, which indicates poor autonomous driving capability. In some implementations, the weight can be set for the value of the automatic driving confidence information in combination with the driving difficulty. Specifically, when the autonomous driving difficulty of the one-way mapping information is high, the first weight can be set for the autonomous driving confidence information calculated above. When the autonomous driving difficulty of the one-way mapping information is normal difficulty, a second weight can be set for the autonomous driving confidence information. When the autonomous driving difficulty of the one-way mapping information is low, a third weight can be set for the autonomous driving confidence information. Wherein, the first weight value is greater than the second weight value and is greater than the third weight value. When the autonomous driving confidence information is 100, there is no need to set the weight. In this way, the difficulty of autonomous driving can be applied to the confidence information of autonomous driving, so that the safety of the vehicle performing autonomous driving can be characterized more safely and accurately.
在本实施方式中，可以根据自动驾驶信心信息，为单程建图信息设置相应的颜色。具体的，在车辆的终端通过语义地图展示该单程建图信息表示的道路时，可以以该颜色表示该单程建图信息的自动驾驶信心信息。具体的，例如，针对前述自动驾驶信心信息为优秀的单程建图信息，颜色可以为墨绿，自动驾驶信心信息为良好的单程建图信息，颜色可以为绿色，自动驾驶信心信息为正常的单程建图信息，颜色可以为黄色，自动驾驶信心信息为差的单程建图信息，颜色可以为红色。In this implementation, corresponding colors can be set for the one-way mapping information based on the autonomous driving confidence information. Specifically, when the vehicle's terminal displays the road represented by the one-way mapping information through a semantic map, the color can be used to represent the autonomous driving confidence information of the one-way mapping information. Specifically, for example, if the aforementioned self-driving confidence information is excellent one-way mapping information, the color can be dark green, if the self-driving confidence information is good one-way mapping information, the color can be green, and if the self-driving confidence information is normal one-way mapping information, the color can be green. The color of the map information can be yellow, and the autonomous driving confidence information is poor one-way mapping information, and the color can be red.
在一些实施方式中，影子模式可以具有验证周期。具体的，验证周期的时长可以为一周。当然，验证周期的时长也可以为10天、15天、或者3天等等。车辆控制系统可以根据验证周期的最后两天的虚拟驾驶数据，生成导航路径信息的自动驾驶信心信息。如此，使得自动驾驶信心信息可以很好的表示车辆的自动驾驶能力。当然，车辆也可以通过显示界面提供设置功能，驾驶员可以设置选择，让车辆控制系统学习自身的驾驶风格，也可以选择不让车辆控制系统学习自身的驾驶风格。In some implementations, shadow mode may have a verification period. Specifically, the length of the verification cycle can be one week. Of course, the length of the verification cycle can also be 10 days, 15 days, or 3 days, etc. The vehicle control system can generate autonomous driving confidence information of navigation path information based on the virtual driving data of the last two days of the verification cycle. In this way, the self-driving confidence information can well represent the vehicle's self-driving capability. Of course, the vehicle can also provide setting functions through the display interface. The driver can set options to let the vehicle control system learn his or her own driving style, or he can choose not to let the vehicle control system learn his or her own driving style.
请参阅图3。自动驾驶路径推荐阶段可以包括以下步骤。See Figure 3. The autonomous driving route recommendation phase may include the following steps.
步骤S31：在驾驶员驾驶车辆的过程中，根据车辆的当前定位数据在熟路组中匹配，得出与所述当前定位数据适配的单程建图信息。Step S31: While the driver is driving the vehicle, the vehicle's current positioning data is matched in the familiar road group to obtain one-way mapping information adapted to the current positioning data.
在本实施方式中，车辆控制系统可以根据车辆的当前定位数据，判断是否行驶在熟路组中单程建图信息对应的道路。在认定车辆行驶于某个单程建图信息对应的道路时，可以将该单程建图信息作为目标单程建图信息。当然，目标单程建图信息表示的道路需要涵盖车辆的当前位置。In this embodiment, the vehicle control system can determine whether the vehicle is traveling on a road corresponding to the one-way mapping information in the familiar road group based on the current positioning data of the vehicle. When it is determined that the vehicle is driving on a road corresponding to a certain one-way mapping information, the one-way mapping information can be used as the target one-way mapping information. Of course, the road represented by the target one-way mapping information needs to cover the current location of the vehicle.
在本实施方式中，在车辆控制系统得出目标单程建图信息之后，可以提醒驾驶员，车辆控制系统已经对当前道路匹配得出目标单程建图信息，使得车辆控制系统已经具备了一定的自动驾驶能力。具体的，可以通过语音的方式对驾驶员进行提醒。In this embodiment, after the vehicle control system obtains the target one-way mapping information, the driver can be reminded that the vehicle control system has matched the current road and obtained the target one-way mapping information, so that the vehicle control system has certain automatic capabilities. driving ability. Specifically, the driver can be reminded through voice.
步骤S32：在所述车辆的终端界面显示目标导航路径信息的自动驾驶信心信息。Step S32: Display the automatic driving confidence information of the target navigation path information on the terminal interface of the vehicle.
在本实施方式中，车辆控制系统可以在车辆的终端界面显示普通导航地图，并显示目标导航路径信息的自动驾驶信心信息。当然，也可以显示所建立的语义地图，以及展示自动驾驶信心信息。具体的，例如，可以直接在终端界面显示自动驾驶信心信息的取值。当然，也可以显示自动驾驶信心信息对应的颜色。具体的，例如，在终端界面显示的普通导航地图中，对应目标单程建图信息表示的道路颜色为目标单程建图信息的自动驾驶信心信息对应的颜色。In this embodiment, the vehicle control system can display a common navigation map on the vehicle's terminal interface, and display the automatic driving confidence information of the target navigation path information. Of course, the established semantic map can also be displayed and autonomous driving confidence information can be displayed. Specifically, for example, the value of the autonomous driving confidence information can be displayed directly on the terminal interface. Of course, the color corresponding to the autonomous driving confidence information can also be displayed. Specifically, for example, in the ordinary navigation map displayed on the terminal interface, the color of the road corresponding to the target one-way mapping information is the color corresponding to the automatic driving confidence information of the target one-way mapping information.
在一些情况下，当目标单程建图信息的自动驾驶信心信息表示的自动驾驶能力，为正常或差时，可以通过语音提醒驾驶员，如果启动车辆的自动驾驶，需要多注意。In some cases, when the automatic driving capability represented by the automatic driving confidence information of the target one-way mapping information is normal or poor, the driver can be reminded through voice that more attention is needed if the automatic driving of the vehicle is started.
步骤S33：根据接收的驾驶员的指令，启动自动驾驶。Step S33: Start automatic driving according to the received driver's instruction.
在本实施方式中，车辆控制系统可以通过语音或者显示界面的按钮，与驾驶员进行交互，在得到驾驶员发出的确定启动自动驾驶指令的情况下，启动车辆的自动驾驶。In this embodiment, the vehicle control system can interact with the driver through voice or buttons on the display interface, and start the vehicle's automatic driving after receiving an instruction from the driver to determine whether to start automatic driving.
请参阅图4。自动驾驶阶段可以包括以下步骤。See Figure 4. The autonomous driving phase can include the following steps.
步骤S41：在车辆行驶过程中，结合车辆的卫星定位数据和惯导定位数据，得到车辆的当前定位数据。Step S41: While the vehicle is driving, the vehicle's current positioning data is obtained by combining the vehicle's satellite positioning data and inertial navigation positioning data.
在本实施方式中，车辆控制系统可以在车辆行驶过程中，不断结合车辆的卫星定位数据和惯导定位数据，以得到由补全的卫星定位数据形成的当前定位数据，具体的，可以参照前述实施方式介绍，不再赘述。In this embodiment, the vehicle control system can continuously combine the vehicle's satellite positioning data and inertial navigation positioning data during the driving process of the vehicle to obtain the current positioning data formed by the completed satellite positioning data. For details, please refer to the above The implementation is introduced and will not be described again.
步骤S42：基于采集到的当前环境数据，进行障碍物和交通标识识别，得到表示障碍物的障碍物数据，和表示交通标识的交通标识数据。Step S42: Based on the collected current environmental data, identify obstacles and traffic signs to obtain obstacle data representing obstacles and traffic sign data representing traffic signs.
在本实施方式中，具体的，可以基于当前环境数据进行障碍物识别、空中交通标识识别、地面交通标识识别等。具体的，可以参见前述实施方式对照解释，不再赘述。In this embodiment, specifically, obstacle recognition, air traffic sign recognition, ground traffic sign recognition, etc. can be performed based on current environmental data. For details, please refer to the foregoing embodiments for comparative explanation, and no further description will be given.
步骤S43：根据所述车辆的当前定位数据，在存储的语义地图中读取指定位置范围内的局部语义地图。Step S43: According to the current positioning data of the vehicle, read the local semantic map within the specified location range from the stored semantic map.
在本实施方式中，车辆控制系统的语义地图引擎可以根据当前定位数据，从语义地图中读取局部语义地图。具体的，可以参见前述实施方式对照解释，不再赘述。In this embodiment, the semantic map engine of the vehicle control system can read the local semantic map from the semantic map according to the current positioning data. For details, please refer to the foregoing embodiments for comparative explanation, and no further description will be given.
步骤S44：结合所述局部语义地图、当前环境数据和当前定位数据，生成车辆的当前融合行驶轨迹。Step S44: Combine the local semantic map, current environment data and current positioning data to generate the current fused driving trajectory of the vehicle.
在本实施方式中，可以将车辆行驶过程中，采集到的当前环境数据和当前定位数据，与局部语义地图进行融合处理，生成车辆的当前融合行驶轨迹。具体的，可以参照前述实施方式对照解释，不再赘述。In this embodiment, the current environment data and current positioning data collected during the driving of the vehicle can be fused with the local semantic map to generate the current fused driving trajectory of the vehicle. Specifically, reference may be made to the foregoing embodiments for comparative explanation, and no further description will be given.
步骤S45：基于从当前环境数据中识别出的地面交通标识数据，进行局部语义建图，得到局部地图。Step S45: Based on the ground traffic sign data identified from the current environmental data, perform local semantic mapping to obtain a local map.
在本实施方式中，可以根据从当前环境数据中识别出的地面交通标识数据进行局部语义建图，得到局部地图。具体的，可以参照前述实施方式对照解释，不再赘述。In this embodiment, local semantic mapping can be performed based on the ground traffic sign data identified from the current environmental data to obtain a local map. Specifically, reference may be made to the foregoing embodiments for comparative explanation, and no further description will be given.
步骤S46：根据表示障碍物的障碍物数据和所述局部地图进行轨迹预测，得到轨迹预测结果。Step S46: Perform trajectory prediction based on the obstacle data representing obstacles and the local map to obtain a trajectory prediction result.
在本实施方式中，轨迹预测模块可以根据表示障碍物的障碍物数据、局部地图进行轨迹预测，以便于自动驾驶决策规划模块进行最终的行为规划和运动规划。具体的，可以参照前述实施方式对照解释，不再赘述。In this embodiment, the trajectory prediction module can perform trajectory prediction based on obstacle data representing obstacles and local maps, so that the autonomous driving decision-making and planning module can perform final behavior planning and motion planning. Specifically, reference may be made to the foregoing embodiments for comparative explanation, and no further description will be given.
步骤S47：基于所述融合行驶轨迹数据、所述局部语义地图的道路对象、所述车辆的导航路径信息和所述轨迹预测结果，进行所述车辆的行为规划和运动规划，以根据所述行为规划和所述运动规划控制所述车辆行驶。Step S47: Based on the fused driving trajectory data, the road objects of the local semantic map, the navigation path information of the vehicle and the trajectory prediction results, perform behavior planning and motion planning of the vehicle to according to the behavior The planning and motion planning control the vehicle travel.
在本实施方式中，车辆控制系统的自动驾驶决策规划模块可以根据生成的行为规划和运动规划，控制车辆实现自动驾驶。In this embodiment, the automatic driving decision planning module of the vehicle control system can control the vehicle to realize automatic driving based on the generated behavior plan and motion plan.
本说明书的一个实施方式还提供一种车辆控制装置。如图5所示，所述车辆控制装置可以包括以下模块。An embodiment of this specification also provides a vehicle control device. As shown in Figure 5, the vehicle control device may include the following modules.
存储模块，用于存储车辆行驶过程中，采集到的车辆的定位数据和环境数据。The storage module is used to store the positioning data and environmental data of the vehicle collected during the driving of the vehicle.
识别模块，用于从存储的环境数据识别出对应定位数据的交通标识信息。The identification module is used to identify the traffic sign information corresponding to the positioning data from the stored environmental data.
建立模块，用于结合交通标识信息和定位数据，建立并存储定位数据对应的指定道路的语义地图，语义地图用于车辆在指定道路执行自动驾驶。The establishment module is used to combine traffic sign information and positioning data to create and store a semantic map of the designated road corresponding to the positioning data. The semantic map is used for the vehicle to perform automatic driving on the designated road.
关于车辆控制装置实现的具体功能和效果，可以参照本说明书其他实施方式对照解释，在此不再赘述。所述车辆控制装置中的各个模块可全部或部分通过软件、硬件及其组合来实现。所述各模块可以硬件形式内嵌于或独立于计算机设备中的处理器中，也可以以软件形式存储于计算机设备中的存储器中，以便于处理器调用执行以上各个模块对应的操作。The specific functions and effects achieved by the vehicle control device can be explained with reference to other embodiments of this specification, and will not be described again here. Each module in the vehicle control device may be implemented in whole or in part by software, hardware and combinations thereof. Each of the modules may be embedded in or independent of the processor of the computer device in the form of hardware, or may be stored in the memory of the computer device in the form of software, so that the processor can call and execute operations corresponding to each of the above modules.
请参阅图6，本说明书实施方式还提供一种车辆的控制方法。所述车辆的控制方法可以应用于车辆。所述车辆的控制方法可以包括以下步骤。Referring to FIG. 6 , the embodiment of this specification also provides a vehicle control method. The vehicle control method may be applied to the vehicle. The vehicle control method may include the following steps.
步骤S50：在所述车辆从第一位置行驶至第二位置的过程中，获取定位数据和环境数据；其中，所述定位数据包括用于表示所述行驶过程中某一时刻的所述车辆的位置的信息，所述环境数据包括用于表示所述行驶过程中某一时刻的所述车辆的周围环境的信息；存储所述定位数据和所述环境数据。Step S50: While the vehicle is traveling from the first position to the second location, obtain positioning data and environmental data; wherein the positioning data includes information representing the vehicle at a certain moment during the driving process. Location information, the environment data includes information representing the surrounding environment of the vehicle at a certain moment during the driving process; the positioning data and the environment data are stored.
在本实施方式中，第一位置可以为车辆一次行程的起点位置。第二位置可以为一次行程的终点位置。车辆从第一位置行驶至第二位置通过车辆的电子控制单元，生成定位数据和环境数据。车辆的车辆控制系统可以不断的对应存储定位数据和环境数据。In this embodiment, the first position may be the starting position of a vehicle trip. The second position may be the end position of a stroke. When the vehicle travels from the first position to the second position, the electronic control unit of the vehicle generates positioning data and environmental data. The vehicle's vehicle control system can continuously store positioning data and environmental data accordingly.
步骤S51：基于所述环境数据，获取与所述车辆的位置相关的交通标识信息。Step S51: Based on the environmental data, obtain traffic sign information related to the location of the vehicle.
在本实施方式中，车辆控制系统可以从环境数据中识别出交通标识信息。具体的，从环境数据识别出交通标识信息，可以参照前述实施方式对照解释，不再赘述。In this embodiment, the vehicle control system can identify traffic sign information from environmental data. Specifically, identification of traffic sign information from environmental data can be explained with reference to the foregoing embodiments, and will not be described again.
步骤S52：至少基于所述交通标识信息，获取第一自建地图信息。Step S52: Obtain first self-built map information based on at least the traffic sign information.
在本实施方式中，将从环境数据中识别的得到的交通标识信息，与按照采集时间对应的定位数据，整体形成第一自建地图信息。在一些实施方式中，第一自建地图信息可以为前文实施方式介绍的单程建图信息。当然，所属领域技术人员在本说明书的实施方式启示下，还可以采用其他实施方式形成所述第一自建地图信息，不再赘述。In this embodiment, the traffic sign information identified from the environmental data and the positioning data corresponding to the collection time are integrated to form the first self-built map information. In some embodiments, the first self-constructed map information may be the one-way mapping information introduced in the previous embodiments. Of course, those skilled in the art, inspired by the implementation of this specification, can also use other implementations to form the first self-built map information, which will not be described again.
步骤S53：根据多于第一阈值的次数的行驶对应的第一自建地图信息，获取第二自建地图信息；其中，所述多于第一阈值的次数的行驶中的每一次行驶的路径包括至少一个相同的路段；其中，所述第二自建地图信息表示地图的精度高于所述第一自建地图信息。Step S53: Obtain the second self-built map information based on the first self-built map information corresponding to the number of trips exceeding the first threshold; wherein the path of each trip in the number of trips exceeding the first threshold is Including at least one identical road section; wherein the second self-built map information indicates that the accuracy of the map is higher than that of the first self-built map information.
在本实施方式中，车辆可以多次行驶于所述第一自建地图信息表示的路段。具体的，例如，驾驶员可能经常驾驶车辆上班和回家。此时，驾驶员每次驾驶车辆上班，便可以形成一个第一自建地图信息，每次回家也可以形成一个第一自建地图信息。In this embodiment, the vehicle can drive on the road section represented by the first self-built map information multiple times. Specifically, for example, a driver may frequently drive a vehicle to work and home. At this time, the driver can generate a first self-built map information every time he drives the vehicle to work, and can also generate a first self-built map information every time he returns home.
在本实施方式中，可以将第一阈值作为一个临界值，判断第一自建地图信息表示的路段是否能够作为熟路。即，可以将被行驶的次数大于第一阈值的第一自建地图信息，认定为熟路，可以被划分熟路组中。可以针对被认定为熟路的第一自建地图信息，对应生成第二自建地图信息。第二自建地图信息可以是车辆控制系统建立的自建地图。具体的，例如，第二自建地图信息可以是车辆控制系统基于语义建图（Simultaneous Localization andMapping，简称SLAM）建立的语义地图。具体的，可以与前述实施方式对照解释。当然，所述第二自建地图信息也可以不限于语义地图，所属领域技术人员可以依照所悉知的技术实现本说明书介绍的技术方案。但只要其实现的功能和效果，与本说明书多个实施方式相同或相似，均应涵盖于本发明保护范围内。In this embodiment, the first threshold can be used as a critical value to determine whether the road section represented by the first self-built map information can be used as a familiar road. That is, the first self-built map information whose number of travel times is greater than the first threshold can be determined as a familiar road and can be classified into a familiar road group. The second self-built map information may be generated correspondingly to the first self-built map information that is recognized as a familiar road. The second self-built map information may be a self-built map created by the vehicle control system. Specifically, for example, the second self-built map information may be a semantic map built by the vehicle control system based on semantic mapping (Simultaneous Localization and Mapping, referred to as SLAM). Specifically, it can be explained in comparison with the foregoing embodiments. Of course, the second self-built map information may not be limited to semantic maps, and those skilled in the art can implement the technical solutions introduced in this specification according to well-known technologies. However, as long as the functions and effects achieved are the same or similar to the multiple embodiments in this specification, they should be covered by the protection scope of the present invention.
在本实施方式中，第一自建地图信息和第二自建地图信息都可以表示车辆控制系统内建立的地图。第二自建地图信息表示的地图的地图精度，可以高于第一自建地图信息的地图精度。通常，生成较高精度的地图，需要更大的运算量，使得需要使用较多的算力资源。在本实施方式中，可以针对被认定为熟路的路算生成较高地图精度的第二自建地图信息，使得建立较高地图精度的路段更加有针对性。即，驾驶员会较多次数行驶于该路段，使得生成的第二自建地图信息也可以具有较多的利用率。In this embodiment, both the first self-built map information and the second self-built map information may represent maps created in the vehicle control system. The map accuracy of the map represented by the second self-built map information may be higher than the map accuracy of the first self-built map information. Generally, generating higher-precision maps requires a greater amount of calculations, which requires the use of more computing resources. In this embodiment, the second self-built map information with higher map accuracy can be generated for the road calculation that is recognized as a familiar road, so that the establishment of road sections with higher map accuracy is more targeted. That is, the driver will drive on this road section more times, so that the generated second self-built map information can also have a greater utilization rate.
在本实施方式中，第一自建地图信息可以为将从环境数据中识别的得到的交通标识信息，与按照采集时间对应的定位数据，整体存储形成。使得，第一自建地图信息可以主要是针对所采集到的定位数据和环境数据的记录。第二自建地图信息可以是基于语义建图构建的语义地图。第二自建地图信息可以为在基于熟路的第一自建地图信息基础上，并进一步的生成第二自建地图信息所需要的数据对象。数据对象可以包括但不限于道路对象和交通标识对象等。再者，第二自建地图信息在构建过程中，可以综合多个第一自建地图信息的数据，使得所生成的数据对象可以具有更加准确的内容和位置。如此，第二自建地图信息可以为针对实际路段的模拟。使得第二自建地图信息的精度高于第一自建地图信息。In this embodiment, the first self-built map information may be formed by storing the traffic sign information identified from the environmental data and the positioning data corresponding to the collection time as a whole. As a result, the first self-built map information may mainly focus on the records of collected positioning data and environmental data. The second self-built map information may be a semantic map constructed based on semantic mapping. The second self-built map information may be based on the first self-built map information based on familiar roads, and further generates data objects required for the second self-built map information. Data objects may include but are not limited to road objects and traffic sign objects. Furthermore, during the construction process of the second self-built map information, the data of multiple first self-built map information can be integrated, so that the generated data object can have more accurate content and location. In this way, the second self-built map information can be a simulation for the actual road section. The accuracy of the second self-built map information is higher than that of the first self-built map information.
步骤S54：存储第二自建地图信息。Step S54: Store the second self-built map information.
在本实施方式中，车辆控制系统可以将获取的第二自建地图信息存储在车辆的存储器中。如此，使得车辆自身完成了针对熟路的地图构建。In this embodiment, the vehicle control system may store the acquired second self-built map information in the memory of the vehicle. In this way, the vehicle itself completes map construction for familiar roads.
步骤S55：至少基于所述第二自建地图信息，控制所述车辆在所述相同的路段执行自动驾驶。Step S55: Based on at least the second self-built map information, control the vehicle to perform automatic driving on the same road section.
在本实施方式中，在车辆再次行驶于熟路的路段时，可以根据第二自建地图信息控制车辆执行自动驾驶。具体的，例如，可以将第二自建地图信息作为车辆控制系统的自动驾驶决策规划模块的部分输入数据，以使得自动驾驶决策规划模块利用第二自建地图信息，输出行为规划和运动规划，进而实现控制车辆在熟路上的自动驾驶。具体的，车辆控制系统基于第二自建地图信息执行自动驾驶的过程，可以与前述实施方式介绍了基于语义地图的自动驾驶的内容对照解释，不再赘述。In this embodiment, when the vehicle drives on a familiar road section again, the vehicle can be controlled to perform automatic driving based on the second self-built map information. Specifically, for example, the second self-built map information can be used as part of the input data of the automatic driving decision planning module of the vehicle control system, so that the automatic driving decision planning module uses the second self-built map information to output behavior planning and motion planning, This enables the vehicle to be controlled for autonomous driving on familiar roads. Specifically, the process of the vehicle control system performing automatic driving based on the second self-built map information can be explained in conjunction with the content of automatic driving based on semantic maps introduced in the previous embodiments, and will not be described again.
在一些实施方式中，所述车辆设置有多个电子控制单元，在获取定位数据和环境数据的步骤中，由所述车辆的多个电子控制单元生成所述定位数据和所述环境数据。In some embodiments, the vehicle is provided with multiple electronic control units, and in the step of obtaining positioning data and environmental data, the positioning data and the environmental data are generated by multiple electronic control units of the vehicle.
在本实施方式中，车辆可以设置有多个电子控制单元，每个电子控制单元可以分别实现车辆所需的功能。电子控制单元也可以与车辆的传感器配合使用，通过传感器采集数据，由电子控制单元进行数据处理。具体的，例如，车辆可以设置有多个环境传感器，并将环境传感器采集到的传感信号提供给相应的电子控制单元，由所述电子控制单元针对传感信号进行处理得出环境数据。车辆还可以设置有生成定位数据的电子控制单元。具体的，电子控制单元可以为定位模组，可以通过GPS系统信号，或者北斗系统信号生成定位数据。电子控制单元还可以包括：惯性导航定位模块。In this embodiment, the vehicle may be provided with multiple electronic control units, and each electronic control unit may separately implement the functions required by the vehicle. The electronic control unit can also be used in conjunction with the vehicle's sensors to collect data through the sensors and the electronic control unit performs data processing. Specifically, for example, the vehicle may be equipped with multiple environmental sensors, and the sensing signals collected by the environmental sensors are provided to the corresponding electronic control unit, and the electronic control unit processes the sensing signals to obtain environmental data. The vehicle may also be provided with an electronic control unit that generates positioning data. Specifically, the electronic control unit can be a positioning module, which can generate positioning data through GPS system signals or Beidou system signals. The electronic control unit may also include: an inertial navigation and positioning module.
在本实施方式中，定位数据和环境数据都在车辆内部署的电子控制单元中运算处理，使得车辆控制系统可以较少的与网络中的服务器交互，提升了车辆控制的自主性。进一步的，车辆控制系统根据自身设置的电子控制单元获取环境数据和定位数据之后，便可以由车辆控制系统自行生成语义地图，而可以无需从服务器获取高精度地图。In this embodiment, both positioning data and environmental data are processed in the electronic control unit deployed in the vehicle, so that the vehicle control system can interact less with the server in the network and improve the autonomy of vehicle control. Furthermore, after the vehicle control system obtains environmental data and positioning data based on its own electronic control unit, the vehicle control system can generate a semantic map on its own without obtaining a high-precision map from the server.
在一些实施方式中，第一自建地图信息对应有作为起点的第一位置，和作为终点的第二位置；针对第一自建地图信息划分熟路组；其中，同一个熟路组中包括的第一自建地图信息之间符合指定关联关系；其中，所述指定关联关系包括：第一自建地图信息的第一位置之间符合第一设定距离条件，且第一自建地图信息的第二位置之间符合第二设定距离条件；或，第一自建地图信息之间涉及的路段的重合度，高于指定重合度阈值。In some implementations, the first self-built map information corresponds to a first location as a starting point and a second location as an end point; familiar road groups are divided for the first self-built map information; wherein, the first self-built map information included in the same familiar road group The self-built map information meets the designated association relationship; wherein, the designated association relationship includes: the first location of the first self-built map information meets the first set distance condition, and the first self-built map information meets the first set distance condition, and the first self-built map information meets the first set distance condition. The second set distance condition is met between the two locations; or, the coincidence degree of the road segments involved in the first self-built map information is higher than the specified coincidence degree threshold.
在本实施方式中，指定关联关系可以包括：单程建图信息的起点信息表示的起点之间符合第一设定距离关系、单程建图信息的终点信息表示的终点之间符合所述第二设定距离关系。当然，第一设定距离关系可以表示起点之间的距离需要满足的条件。第二设定距离关系可以表示终点之间的距离需要满足的条件。第一设定距离关系和第二设定距离关系可以相同，当然，也可以不同。具体的，第一设定距离关系和第二距离设定关系，可以依照实际需求进行设定。具体的，例如，第一设定距离关系为小于200米，第二设定距离关系也可以为小于200米。当然，第一设定距离关系可以为小于100米，第二设定距离关系可以为小于150米。在一些实施方式中，所述指定关联关系也可以包括路径重合度大于指定重合度阈值。具体的，多个路径信息的起点、途经点和终点总体形成路径之间的重合度。指定重合度阈值可以为70%，或75%、80%等等。在一些实施方式中，所述熟路组可以理解为车辆已经至少一次行驶于熟路组中单程建图信息对应的道路。当然，在一些实施方式中，也可以针对车辆行驶于单程建图信息对应道路的次数进行限定，在次数大于指定行驶次数阈值的情况下，才将单程建图信息划分入熟路组。In this embodiment, the specified association relationship may include: the starting points represented by the starting point information of the one-way mapping information comply with the first set distance relationship, and the end points represented by the end point information of the one-way mapping information comply with the second set distance relationship. distance relationship. Of course, the first set distance relationship can represent the conditions that the distance between starting points needs to satisfy. The second set distance relationship can represent the conditions that the distance between the end points needs to meet. The first set distance relationship and the second set distance relationship may be the same, or of course, may be different. Specifically, the first set distance relationship and the second distance set relationship can be set according to actual needs. Specifically, for example, the first set distance relationship is less than 200 meters, and the second set distance relationship can also be less than 200 meters. Of course, the first set distance relationship may be less than 100 meters, and the second set distance relationship may be less than 150 meters. In some embodiments, the specified association relationship may also include that the path coincidence degree is greater than a specified coincidence degree threshold. Specifically, the starting points, passing points and end points of multiple path information collectively form the degree of overlap between paths. The specified coincidence threshold can be 70%, or 75%, 80%, etc. In some embodiments, the familiar road group can be understood as the vehicle has traveled on the road corresponding to the one-way mapping information in the familiar road group at least once. Of course, in some implementations, the number of times a vehicle travels on the road corresponding to the one-way mapping information can also be limited. Only when the number of times the vehicle travels is greater than a specified threshold of the number of trips, the one-way mapping information is divided into the familiar road group.
在一些实施方式中，至少基于所述第二自建地图信息，控制所述车辆在所述相同的路段执行自动驾驶的步骤，包括：基于所述第二自建地图信息，生成所述车辆行驶于路段的融合行驶轨迹；从所述融合行驶轨迹中解析得到所述车辆相对于路段的纵向和/或横向的目标速度分布；调整所述车辆的自动驾驶决策规划模块，以使所述车辆自动驾驶于所述熟路组对应的路段的过程中，所述车辆的速度分布趋于所述目标速度分布。In some embodiments, based on at least the second self-built map information, the step of controlling the vehicle to perform automatic driving on the same road segment includes: based on the second self-built map information, generating the vehicle driving The fused driving trajectory of the road segment; the longitudinal and/or transverse target speed distribution of the vehicle relative to the road segment is obtained from the fused driving trajectory; the automatic driving decision planning module of the vehicle is adjusted to make the vehicle automatically During driving on the road section corresponding to the familiar road group, the speed distribution of the vehicle tends to the target speed distribution.
在本实施方式中，可以基于第二自建地图信息生成车辆行驶过路段的融合行驶轨迹。具体的，第二自建地图信息可以为单程建图信息。在同一个熟路组中的多个单程建图信息，会涉及至少部分相同的路段。至少可以针对该些相同的路段，生成车辆的融合定位轨迹。具体的，建立融合定位轨迹的方式，可以与前述实施方式对照解释，不再赘述。In this embodiment, the integrated driving trajectory of the road section traveled by the vehicle can be generated based on the second self-built map information. Specifically, the second self-built map information may be one-way map building information. Multiple one-way mapping information in the same familiar road group will involve at least some of the same road segments. At least the fused positioning trajectory of the vehicle can be generated for the same road sections. Specifically, the method of establishing the fusion positioning trajectory can be explained in comparison with the previous embodiments, and will not be described again.
在本实施方式中，生成融合行驶轨迹之后，可以从融合行驶轨迹中解析得到车辆相对于路段的纵向和/或横向的目标速度分布。如此，便可以调整车辆的自动驾驶决策规划模块，以使所述车辆自动驾驶于所述熟路组对应的路段的过程中，所述车辆的速度分布趋于所述目标速度分布。在一些实施方式中，可以根据目标速度分布确定驾驶员的驾驶风格，进而可以在调整自动驾驶决策规划模块时，使得自动驾驶决策规划模块可以学习驾驶员的驾驶风格，车辆在熟路上行驶过程中的速度分布趋于目标速度分布。In this embodiment, after the fused driving trajectory is generated, the longitudinal and/or transverse target speed distribution of the vehicle relative to the road segment can be obtained by analyzing the fused driving trajectory. In this way, the automatic driving decision planning module of the vehicle can be adjusted so that when the vehicle automatically drives on the road section corresponding to the familiar road group, the speed distribution of the vehicle tends to the target speed distribution. In some embodiments, the driver's driving style can be determined based on the target speed distribution, and then when the automatic driving decision planning module is adjusted, the automatic driving decision planning module can learn the driver's driving style. When the vehicle is driving on a familiar road The speed distribution tends to the target speed distribution.
在一些实施方式中，基于所述第二自建地图信息，生成所述第一自建地图信息对应的融合行驶轨迹的步骤，可以包括：根据所述车辆的定位数据，在存储的第二自建地图信息中读取指定位置范围内的局部地图；结合所述局部地图、环境数据和定位数据，生成车辆的融合行驶轨迹。In some embodiments, the step of generating the fused driving trajectory corresponding to the first self-built map information based on the second self-built map information may include: based on the positioning data of the vehicle, in the stored second self-built map information. The local map within the specified location range is read from the map information; and the integrated driving trajectory of the vehicle is generated by combining the local map, environmental data and positioning data.
在本实施方式中，在影子模式验证阶段，可以根据车辆的定位数据，在存储的第二自建地图信息中定位，得到局部地图。具体的，可以读取以第二自建地图信息中定位的位置为中心，处于指定位置范围内的局部地图。在一个实施方式中，第二自建地图信息是语义地图。可以根据所述车辆的定位数据，在存储的语义地图中读取指定位置范围内的局部语义地图，结合所述局部语义地图、环境数据和定位数据，生成车辆的融合行驶轨迹。具体的，可以参照前述实施方式对照解释，不再赘述。In this embodiment, during the shadow mode verification stage, the local map can be obtained by positioning in the stored second self-built map information based on the positioning data of the vehicle. Specifically, a local map centered on the position located in the second self-built map information and within the specified position range can be read. In one implementation, the second self-built map information is a semantic map. According to the positioning data of the vehicle, the local semantic map within the specified location range can be read from the stored semantic map, and the integrated driving trajectory of the vehicle can be generated by combining the local semantic map, environmental data and positioning data. Specifically, reference may be made to the foregoing embodiments for comparative explanation, and no further description will be given.
在一些实施方式中，调整所述车辆的自动驾驶决策规划模块，以使所述车辆自动驾驶于所述熟路组对应的路段的过程中，所述车辆的速度分布趋于所述目标速度分布的步骤，可以包括：基于从环境数据中识别出的地面交通标识数据，进行局部建图，得到局部地图；根据从环境数据中识别出的表示障碍物的障碍物数据和所述局部地图进行轨迹预测，得到轨迹预测结果；基于所述融合行驶轨迹数据、所述局部地图的道路对象、所述车辆的导航路径信息和所述轨迹预测结果，进行所述车辆的虚拟行为规划和虚拟运动规划；将基于所述虚拟行为规划和所述虚拟运动规划的虚拟控制数据，与驾驶员的实际控制数据进行比较，得出差异数据，以根据所述差异数据修改所述自动驾驶决策规划模块。In some embodiments, the automatic driving decision planning module of the vehicle is adjusted so that when the vehicle automatically drives on the road section corresponding to the familiar road group, the speed distribution of the vehicle tends to the target speed distribution. The steps may include: performing local mapping based on the ground traffic identification data identified from the environmental data to obtain a local map; and performing trajectory prediction based on the obstacle data representing obstacles identified from the environmental data and the local map. , obtain the trajectory prediction results; based on the fused driving trajectory data, the road objects of the local map, the navigation path information of the vehicle and the trajectory prediction results, perform virtual behavior planning and virtual motion planning of the vehicle; The virtual control data based on the virtual behavior planning and the virtual motion planning is compared with the driver's actual control data to obtain difference data, so as to modify the automatic driving decision planning module according to the difference data.
本实施方式中的内容，可以与影子模式验证阶段相关实施方式的介绍内容对照解释，不再赘述。The content in this implementation mode can be explained in conjunction with the introduction content of the implementation mode related to the shadow mode verification phase, and will not be described again.
在一些实施方式中，所述车辆的控制方法还可以包括：根据所述目标速度分布按照设定难度规则，确定路段对应的驾驶难度；其中，所述驾驶难度包括高难度或低难度；其中，所述设定难度规则包括：所述纵向速度分布表示所述车辆的速度小于指定速度阈值，认定所述路段对应的驾驶难度为高难度；或者，所述纵向速度分布表示所述车辆的速度分布均匀，且平均速度大于所述指定速度阈值，认定所述路段对应的驾驶难度为低难度。In some embodiments, the vehicle control method may further include: determining the driving difficulty corresponding to the road section according to the set difficulty rule according to the target speed distribution; wherein the driving difficulty includes high difficulty or low difficulty; wherein, The setting difficulty rules include: the longitudinal speed distribution indicates that the speed of the vehicle is less than a specified speed threshold, and the driving difficulty corresponding to the road section is determined to be high difficulty; or the longitudinal speed distribution indicates the speed distribution of the vehicle Evenly, and the average speed is greater than the specified speed threshold, the driving difficulty corresponding to the road section is determined to be low difficulty.
在本实施方式中，具体的，例如，如果一个道路的纵向的目标速度分布数据小于一个指定的指定速度阈值，可以判定该道路为经常拥堵路段，可以认定该路为高难度。具体的，指定速度阈值的速度取值可以较小，比如30KM/H，或25KM/H等。如果一个道路的纵向目标速度分布数据较为均匀的维持在一个较快速度，且平均速度大于指定速度阈值，可以认定该道路为低难度。In this embodiment, specifically, for example, if the longitudinal target speed distribution data of a road is less than a specified specified speed threshold, the road can be determined to be a frequently congested road section, and the road can be determined to be highly difficult. Specifically, the speed value of the specified speed threshold can be smaller, such as 30KM/H, or 25KM/H, etc. If the longitudinal target speed distribution data of a road is relatively uniform and maintained at a relatively fast speed, and the average speed is greater than the specified speed threshold, the road can be deemed to be of low difficulty.
在一些实施方式中，所述车辆的控制方法还包括：根据所述差异数据，生成所述第二自建地图信息中路段对应的自动驾驶信心信息；其中，在所述第二自建地图信息中，通过所述路段的颜色，表示所述自动驾驶信心信息。In some embodiments, the vehicle control method further includes: generating automatic driving confidence information corresponding to the road section in the second self-built map information based on the difference data; wherein, in the second self-built map information , the autonomous driving confidence information is represented by the color of the road section.
在本实施方式中，生成路段对应的自动驾驶信心信息，可以与前述介绍的针对单程建图信息的自动驾驶信心信息的实施方式对照解释。In this embodiment, the automatic driving confidence information corresponding to the road segment is generated, which can be explained in comparison with the implementation of the automatic driving confidence information for one-way mapping information introduced above.
在本实施方式中，可以根据自动驾驶信心信息，为第二自建地图信息中路段设置相应的颜色。具体的，在车辆的终端通过第二自建地图信息展示该路段时，可以以该颜色表示该路段的自动驾驶信心信息。具体的，例如，针对前述自动驾驶信心信息为优秀的路段，颜色可以为墨绿，自动驾驶信心信息为良好的路段，颜色可以为绿色，自动驾驶信心信息为正常的路段，颜色可以为黄色，自动驾驶信心信息为差的路段，颜色可以为红色。In this embodiment, corresponding colors can be set for the road sections in the second self-built map information based on the autonomous driving confidence information. Specifically, when the vehicle terminal displays the road section through the second self-built map information, the color can be used to represent the automatic driving confidence information of the road section. Specifically, for example, for the road sections where the self-driving confidence information is excellent, the color can be dark green, for the road sections where the self-driving confidence information is good, the color can be green, and for the road sections where the self-driving confidence information is normal, the color can be yellow, and the automatic Road sections with poor driving confidence information can be colored red.
请参阅图7。本说明书实施方式还提供一种自动驾驶提示方法。所述方法应用于所述车辆。所述自动驾驶提示方法可以包括以下步骤。See Figure 7. The embodiment of this specification also provides an automatic driving prompt method. The method is applied to the vehicle. The automatic driving prompt method may include the following steps.
步骤S60：确定所述车辆从指定的第一位置行驶至指定的第二位置涉及的路段。Step S60: Determine the road section involved in the vehicle traveling from the designated first location to the designated second location.
步骤S61：获取所述路段中至少部分子路段的自动驾驶信心信息；其中，所述自动驾驶信心信息为根据子路段对应的历史差异数据生成，所述历史差异数据是基于执行该子路段的自动驾驶决策算法生成的虚拟控制数据与驾驶员驾驶所述车辆驶过所述子路段执行的实际控制数据获得的；所述虚拟控制数据是由所述自动驾驶决策算法将根据自建地图信息、环境数据和定位数据进行处理，得到所述车辆的车道级定位的目标融合行驶轨迹，并基于所述目标融合行驶轨迹而生成的；其中，所述自建地图信息为所述车辆构建的地图信息。Step S61: Obtain the automatic driving confidence information of at least some sub-sections in the road section; wherein the automatic driving confidence information is generated based on the historical difference data corresponding to the sub-section, and the historical difference data is based on the execution of the automatic driving of the sub-section. The virtual control data generated by the driving decision-making algorithm is obtained from the actual control data performed by the driver driving the vehicle through the sub-section; the virtual control data is obtained by the automatic driving decision-making algorithm based on the self-built map information and environment. The data and positioning data are processed to obtain the target fusion driving trajectory of the vehicle's lane-level positioning, and the target fusion driving trajectory is generated based on the target fusion driving trajectory; wherein the self-built map information is the map information constructed by the vehicle.
步骤S62：提示所述自动驾驶信心信息。Step S62: Prompt the automatic driving confidence information.
在本实施方式中，自建地图信息可以为车辆的车辆控制系统自行执行语义建图生成的语义地图。具体的，可以与前述实施方式揭示的第二自建地图信息对照揭示。In this embodiment, the self-built map information may be a semantic map generated by the vehicle control system performing semantic mapping on its own. Specifically, it may be disclosed in comparison with the second self-built map information disclosed in the foregoing embodiments.
在本实施方式中，第一位置可以用于表示行程的起点位置。第二位置可以用于表示行程的终点位置。在一些实施方式中，第一位置可以为车辆控制系统采集的车辆当前的定位数据表示的位置。第二位置可以是驾驶员向车辆控制系统输入的目标位置。从第一位置至目标位置可以形成一个路段。所述路段中可以包括多个子路段。具体的，例如，路段中可以有途经点。如此，第一位置至最近的途经点，相邻途经点之间，或，最后的途经点与第二位置之间可以分别形成子路段。当然，在一些实施方式中，也可以将路段中涉及的每个道路作为一个子路段。In this embodiment, the first position may be used to represent the starting point of the stroke. The second position can be used to represent the end position of the stroke. In some embodiments, the first position may be a position represented by the current positioning data of the vehicle collected by the vehicle control system. The second position may be a target position input by the driver to the vehicle control system. A road segment can be formed from the first position to the target position. The road segment may include multiple sub-road segments. Specifically, for example, there may be passing points in the road segment. In this way, sub-sections can be formed between the first location and the nearest way point, between adjacent way points, or between the last way point and the second location. Of course, in some implementations, each road involved in the road segment can also be regarded as a sub-road segment.
在本实施方式中，可以分别将车辆历史上行驶于子路段时计算得出的差异数据，作为每个子路段的历史差异数据。进而，可以根据前文实施方式介绍的生成每个子路段的自动驾驶信心信息。车辆的控制系统可以在驾驶员驾驶的过程中，至少基于环境信息执行自动驾驶决策算法的运算。如此，车辆控制系统会模拟自动驾驶车辆的控制过程，生成虚拟控制数据。如此，将虚拟控制数据与车辆被驾驶员驾驶产生的实际控制数据比较，可以得出表示车辆控制系统控制车辆执行自动驾驶的结果，与驾驶员驾驶车辆的差异程度的差异数据。如此，车辆控制系统可以动态的修改自动驾驶决策算法中的参数，以期望减少差异数据表示的差异程度，使得车辆控制系统可以学习驾驶员驾驶车辆的驾驶习惯。In this embodiment, the difference data calculated when the vehicle historically traveled on the sub-section can be used as the historical difference data of each sub-section. Furthermore, the automatic driving confidence information of each sub-section can be generated according to the previous embodiments. The vehicle's control system can perform the operation of the automatic driving decision-making algorithm based on at least environmental information while the driver is driving. In this way, the vehicle control system will simulate the control process of the autonomous vehicle and generate virtual control data. In this way, by comparing the virtual control data with the actual control data generated when the vehicle is driven by the driver, difference data can be obtained that represents the difference between the result of the vehicle control system controlling the vehicle to perform automatic driving and the difference between the vehicle driven by the driver. In this way, the vehicle control system can dynamically modify the parameters in the automatic driving decision algorithm in order to reduce the degree of difference represented by the difference data, so that the vehicle control system can learn the driver's driving habits of the vehicle.
在本实施方式中，自动驾驶决策算法可以集成于车辆控制系统中。具体的，例如，车辆控制系统中可以具有自动驾驶决策规划模块，自动驾驶决策算法可以应用于自动驾驶决策规划模块。In this embodiment, the automatic driving decision algorithm can be integrated into the vehicle control system. Specifically, for example, the vehicle control system may have an automatic driving decision planning module, and the automatic driving decision algorithm may be applied to the automatic driving decision planning module.
在一些实施方式中，车辆控制系统也可以在车辆行驶于某一路段的次数多于指定行驶次数阈值的情况下，才使用存储的定位数据和环境数据基于自建地图信息，生成目标融合行驶轨迹，进而可以根据目标融合行驶轨迹基于自动驾驶决策算法生成虚拟控制数据。并将虚拟控制数据与存储的驾驶员的实际控制数据进行比较得出历史差异数据。实现，对应子路段生成自动驾驶信心信息。如此，在生成自动驾驶信心信息之后，可以存储在车辆设置的存储器中，以便于读取使用。自动驾驶信心信息可以随着车辆进一步行驶于相应子路段进行更新。In some embodiments, the vehicle control system can also use the stored positioning data and environmental data to generate a target fusion driving trajectory based on self-built map information only when the number of times the vehicle travels on a certain road section is greater than the specified driving number threshold. , and then virtual control data can be generated based on the automatic driving decision-making algorithm according to the target fusion driving trajectory. The virtual control data is compared with the stored driver's actual control data to obtain historical difference data. Implementation, generating automatic driving confidence information corresponding to sub-sections. In this way, after the automatic driving confidence information is generated, it can be stored in the memory of the vehicle settings for easy reading and use. The autonomous driving confidence information can be updated as the vehicle further drives on the corresponding sub-section.
本实施方式涉及的技术方案的细节，可以参照前述影子模式验证阶段中相关实施方式对照解释，不再赘述。The details of the technical solutions involved in this implementation can be explained with reference to the related implementations in the aforementioned shadow mode verification stage, and will not be described again.
在本实施方式中，车辆控制系统可以控制车辆向驾驶员提示所行驶的路段对应的自动驾驶信心信息。如此，驾驶员可以自行判断是否启动自动驾驶。In this embodiment, the vehicle control system can control the vehicle to prompt the driver with automatic driving confidence information corresponding to the road section being traveled. In this way, the driver can decide whether to activate autonomous driving.
本实施方式涉及的技术方案的细节，可以参照前述自动驾驶路径推荐阶段中相关实施方式对照解释，不再赘述。The details of the technical solution involved in this implementation mode can be explained with reference to the related implementation modes in the aforementioned autonomous driving route recommendation stage, and will not be described again.
在一些实施方式中，提示所述自动驾驶信心信息的步骤，可以包括：控制所述车辆的车载显示器，提示所述车辆能执行自动驾驶的路段，以及对应路段提示路段的自动驾驶信心信息。In some embodiments, the step of prompting the automatic driving confidence information may include: controlling an on-board display of the vehicle, prompting the road section where the vehicle can perform automatic driving, and prompting the automatic driving confidence information of the corresponding road section.
在本实施方式中，车辆控制系统可以控制车辆的车载显示器中，可以提示能够执行自动驾驶的路段，以及路段的自动驾驶信心信息。如此，驾驶员可以快速的了解车辆对于所行驶路段或后续路段中，车辆的自动驾驶能力。便于驾驶员决定是否启动自动驾驶，或者，便于驾驶员规划在后续路段启动自动驾驶。具体的，例如，车辆控制系统可以控制车载显示器显示普通导航地图，并可以针对普通导航地图中路段的颜色表示自动驾驶信心信息。如此，驾驶员看到相关界面之后，便可以了解路段能够执行自动驾驶，以及车辆执行自动驾驶的能力情况。In this embodiment, the vehicle control system can control the vehicle's on-board display to prompt the road sections where autonomous driving can be performed, as well as the autonomous driving confidence information of the road sections. In this way, the driver can quickly understand the vehicle's autonomous driving capabilities for the road section it is driving or subsequent road sections. It is convenient for the driver to decide whether to start automatic driving, or it is convenient for the driver to plan to start automatic driving on subsequent road sections. Specifically, for example, the vehicle control system can control the on-board display to display a common navigation map, and can express autonomous driving confidence information based on the color of the road section in the common navigation map. In this way, after the driver sees the relevant interface, he can understand that the road section can perform automatic driving and the vehicle's ability to perform automatic driving.
本实施方式涉及的技术方案的细节，可以参照前述自动驾驶路径推荐阶段中相关实施方式对照解释，不再赘述。The details of the technical solution involved in this implementation mode can be explained with reference to the related implementation modes in the aforementioned autonomous driving route recommendation stage, and will not be described again.
在一些实施方式中，控制所述车辆的车载显示器，提示所述车辆能执行自动驾驶的路段，以及对应路段提示路段的自动驾驶信心信息的步骤，可以包括：区分不同的路段，分别提示每个路段的自动驾驶信心信息。In some embodiments, the step of controlling the on-board display of the vehicle to prompt the road sections on which the vehicle can perform autonomous driving, and prompting the self-driving confidence information of the corresponding road sections, may include: distinguishing different road sections and prompting each road section respectively. Autonomous driving confidence information for road sections.
在本实施方式中，不同路段的自动驾驶信心信息之间相互独立。即，不同路段的自动驾驶信心信息可以相同，也可以不相同。如此，分别提示每个路段的自动驾驶信心信息，可以较为准确的表达车辆控制系统于不同路段的自动驾驶能力。如此，给了驾驶员较为准确信息反馈，以便于驾驶员在不同路段具体采用较为适当的驾驶行为。具体的，例如，针对自动驾驶信心信息表示自动驾驶能力为差的路段，驾驶员在启动自动驾驶功能之后，也需要谨慎观察外部情况，以便于出现意外情况即时接管车辆。在本实施方式中，具体的，例如，为了较为明显的区分不同的路段，在车载显示器中可以通过颜色，或者，纹路，图案来提示相同或者不同的自动驾驶信心信息。In this implementation, the autonomous driving confidence information of different road sections is independent of each other. That is, the autonomous driving confidence information of different road sections may be the same or different. In this way, the automatic driving confidence information of each road section is prompted separately, which can more accurately express the automatic driving capabilities of the vehicle control system in different road sections. In this way, the driver is given more accurate information feedback, so that the driver can adopt more appropriate driving behaviors on different road sections. Specifically, for example, for road sections where the self-driving confidence information indicates that the self-driving capability is poor, the driver also needs to carefully observe the external situation after activating the self-driving function in order to take over the vehicle immediately if an unexpected situation occurs. In this embodiment, specifically, for example, in order to more clearly distinguish different road sections, the same or different autonomous driving confidence information can be prompted on the vehicle-mounted display through colors, textures, and patterns.
本实施方式涉及的技术方案的细节，可以参照前述自动驾驶路径推荐阶段中相关实施方式对照解释，不再赘述。The details of the technical solution involved in this implementation mode can be explained with reference to the related implementation modes in the aforementioned autonomous driving route recommendation stage, and will not be described again.
在一些实施方式中，向驾驶员反馈所述自动驾驶信心信息表示的执行自动驾驶的驾驶能力的步骤，可以包括：控制所述车辆的车载显示器，根据自动驾驶信心信息区分颜色展示路段；其中，路段的颜色用于表示所述车辆执行自动驾驶的驾驶能力；或者，通过语音播报，所述路段的自动驾驶信心信息表示的执行自动驾驶的驾驶能力。In some embodiments, the step of feeding back to the driver the driving ability to perform autonomous driving represented by the autonomous driving confidence information may include: controlling an on-board display of the vehicle and displaying road sections in different colors according to the autonomous driving confidence information; wherein, The color of the road section is used to represent the driving ability of the vehicle to perform automatic driving; or, through voice broadcast, the automatic driving confidence information of the road section represents the driving ability to perform automatic driving.
在本实施方式中，可以通过不同的颜色表示不同路段的自动驾驶的驾驶能力。驾驶员可以在查看车载显示器之后，了解所处路段的自动驾驶的驾驶能力。在一些实施方式中，车辆的控制系统也可以通过语音播报的方式，告知驾驶员车辆所处路段的自动驾驶的驾驶能力。如此，驾驶员可以无需查看车载显示器，便可以获知车辆所处路段的自动驾驶的驾驶能力。In this embodiment, different colors can be used to represent the driving capabilities of autonomous driving on different road sections. Drivers can check the on-board display to understand the autonomous driving capabilities of the road section they are on. In some embodiments, the vehicle's control system can also inform the driver of the autonomous driving capability of the road section where the vehicle is located through voice broadcast. In this way, the driver can learn the autonomous driving capabilities of the road section where the vehicle is located without looking at the on-board display.
在本实施方式中，车载显示器可以为设置在车辆内的电子装置。具体的，车载显示器可以为应用于车辆的LCD或LED显示器。当然，在一些实施方式中，具体的，例如，车载显示器可以为汽车平视显示器（Head Up Display，HUD），或，增强现实显示设备（AugmentedReality，AR）等。In this embodiment, the vehicle-mounted display may be an electronic device installed in the vehicle. Specifically, the vehicle-mounted display may be an LCD or LED display applied to the vehicle. Of course, in some implementations, specifically, for example, the vehicle-mounted display may be a car head-up display (HUD), or an augmented reality display device (AugmentedReality, AR), etc.
请参阅图8。本说明书实施方式提供一种车辆的控制装置。所述车辆的控制装置包括以下模块。See Figure 8. An embodiment of this specification provides a vehicle control device. The vehicle control device includes the following modules.
第一获取模块，用于在所述车辆从第一位置行驶至第二位置的过程中，获取定位数据和环境数据；其中，所述定位数据包括用于表示所述行驶过程中某一时刻的所述车辆的位置的信息，所述环境数据包括用于表示所述行驶过程中某一时刻的所述车辆的周围环境的信息；存储所述定位数据和所述环境数据。The first acquisition module is used to acquire positioning data and environmental data while the vehicle is traveling from the first position to the second position; wherein the positioning data includes a location data representing a certain moment in the driving process. Information about the location of the vehicle, and the environmental data includes information representing the surrounding environment of the vehicle at a certain moment during the driving process; the positioning data and the environmental data are stored.
第二获取模块，用于基于所述环境数据，获取与所述车辆的位置相关的交通标识信息。A second acquisition module is configured to acquire traffic sign information related to the location of the vehicle based on the environmental data.
第三获取模块，用于至少基于所述交通标识信息，获取第一自建地图信息。The third acquisition module is used to acquire the first self-built map information based on at least the traffic sign information.
第四获取模块，用于根据多于第一阈值的次数的行驶对应的第一自建地图信息，获取第二自建地图信息；其中，所述多于第一阈值的次数的行驶中的每一次行驶的路径包括至少一个相同的路段。The fourth acquisition module is configured to obtain the second self-built map information based on the first self-built map information corresponding to the number of trips exceeding the first threshold; wherein each of the trips exceeding the first threshold is A route traveled at one time includes at least one identical road segment.
存储模块，用于存储第二自建地图信息。The storage module is used to store the second self-built map information.
控制模块，用于至少基于所述第二自建地图信息，控制所述车辆在所述相同的路段执行自动驾驶。A control module configured to control the vehicle to perform automatic driving on the same road section based on at least the second self-built map information.
在本实施方式中，所述车辆的控制装置实现的功能和效果，可以与前述实施方式对照解释，不再赘述。In this embodiment, the functions and effects achieved by the vehicle control device can be explained in comparison with the previous embodiments, and will not be described again.
请参阅图9。本说明书实施方式提供一种自动驾驶提示装置，所述装置包括：确定模块，用于确定所述车辆从指定的第一位置行驶至指定的第二位置涉及的路段；信心信息获取模块，用于获取所述路段中至少部分子路段的自动驾驶信心信息；其中，所述自动驾驶信心信息为根据子路段对应的历史差异数据生成，所述历史差异数据是基于执行该子路段的自动驾驶决策算法生成的虚拟控制数据与驾驶员驾驶所述车辆驶过所述子路段执行的实际控制数据获得的；所述虚拟控制数据是由所述自动驾驶决策算法将根据自建地图信息、环境数据和定位数据进行处理，得到所述车辆的车道级定位的目标融合行驶轨迹，并基于所述目标融合行驶轨迹而生成的；其中，所述自建地图信息为所述车辆构建的地图信息；提示模块，用于提示所述自动驾驶信心信息。See Figure 9. The embodiment of this specification provides an automatic driving prompt device, which includes: a determination module for determining the road section involved in the vehicle traveling from a designated first position to a designated second position; and a confidence information acquisition module for Obtain the automatic driving confidence information of at least part of the sub-sections in the road section; wherein the automatic driving confidence information is generated based on the historical difference data corresponding to the sub-section, and the historical difference data is based on the execution of the automatic driving decision algorithm of the sub-section. The generated virtual control data is obtained from the actual control data performed by the driver driving the vehicle through the sub-section; the virtual control data is generated by the automatic driving decision-making algorithm based on self-built map information, environmental data and positioning The data is processed to obtain the target fusion driving trajectory of the vehicle's lane-level positioning, and is generated based on the target fusion driving trajectory; wherein the self-built map information is the map information constructed by the vehicle; the prompt module, Used to prompt the automatic driving confidence information.
在本实施方式中，所述自动驾驶提示装置实现的功能和效果，可以与前述实施方式对照解释，不再赘述。In this embodiment, the functions and effects achieved by the automatic driving prompt device can be explained in comparison with the previous embodiments, and will not be described again.
请参阅图10。本说明书实施方式还提供一种电子设备，包括存储器和处理器，所述存储器存储有计算机程序，其特征在于，所述处理器执行所述计算机程序时实现上述任一实施方式中的车辆控制方法。See Figure 10. An embodiment of this specification also provides an electronic device, including a memory and a processor. The memory stores a computer program. The characteristic is that when the processor executes the computer program, it implements the vehicle control method in any of the above embodiments. .
所述电子设备可以包括被系统总线连接的处理器、非易失性存储介质、内存储器、通信接口、显示装置和输入装置。所述非易失性存储介质可以存储有操作系统和相关的计算机程序。The electronic device may include a processor, a non-volatile storage medium, an internal memory, a communication interface, a display device and an input device connected by a system bus. The non-volatile storage medium may store an operating system and related computer programs.
本说明书实施方式还提供一种计算机可读存储介质，其上存储有计算机程序，该计算机程序被计算机执行时使得，该计算机执行上述任一实施方式中的车辆控制方法。Embodiments of this specification also provide a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a computer, the computer executes the vehicle control method in any of the above embodiments.
可以理解，本文中的具体的例子只是为了帮助本领域技术人员更好地理解本说明书实施方式，而非限制本发明的范围。It can be understood that the specific examples herein are only to help those skilled in the art better understand the embodiments of this specification, but are not intended to limit the scope of the present invention.
可以理解，在本说明书中的各种实施方式中，各过程的序号的大小并不意味着执行顺序的先后，各过程的执行顺序应以其功能和内在逻辑确定，而不应对本说明书实施方式的实施过程构成任何限定。It can be understood that in the various implementations of this specification, the size of the serial number of each process does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not be determined by the implementation of this specification. The implementation process constitutes any limitation.
可以理解，本说明书中描述的各种实施方式，既可以单独实施，也可以组合实施，本说明书实施方式对此并不限定。It can be understood that the various embodiments described in this specification can be implemented individually or in combination, and the embodiments in this specification are not limited to this.
除非另有说明，本说明书实施方式所使用的所有技术和科学术语与本说明书的技术领域的技术人员通常理解的含义相同。本说明书中所使用的术语只是为了描述具体的实施方式的目的，不是旨在限制本说明书的范围。本说明书所使用的术语“和/或”包括一个或多个相关的所列项的任意的和所有的组合。在本说明书实施方式和所附权利要求书中所使用的单数形式的“一种”、“上述”和“该”也旨在包括多数形式，除非上下文清楚地表示其他含义。Unless otherwise stated, all technical and scientific terms used in the embodiments of this specification have the same meanings as commonly understood by those skilled in the technical field of this specification. The terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit the scope of this specification. As used in this specification, the term "and/or" includes any and all combinations of one or more of the associated listed items. As used in this specification and the appended claims, the singular forms "a," "above," and "the" are intended to include the plural forms as well, unless the context clearly dictates otherwise.
可以理解，本说明书实施方式的处理器可以是一种集成电路芯片，具有信号的处理能力。在实现过程中，上述方法实施方式的各步骤可以通过处理器中的硬件的集成逻辑电路或者软件形式的指令完成。上述的处理器可以是通用处理器、数字信号处理器(Digital SignalProcessor，DSP)、专用集成电路(Application Specific IntegratedCircuit，ASIC)、现成可编程门阵列(Field Programmable Gate Array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。可以实现或者执行本说明书实施方式中的公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。结合本说明书实施方式所公开的方法的步骤可以直接体现为硬件译码处理器执行完成，或者用译码处理器中的硬件及软件模块组合执行完成。软件模块可以位于随机存储器，闪存、只读存储器，可编程只读存储器或者电可擦写可编程存储器、寄存器等本领域成熟的存储介质中。该存储介质位于存储器，处理器读取存储器中的信息，结合其硬件完成上述方法的步骤。It can be understood that the processor in the embodiment of this specification may be an integrated circuit chip with signal processing capabilities. During the implementation process, each step of the above method implementation may be completed through an integrated logic circuit of hardware in the processor or instructions in the form of software. The above-mentioned processor can be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), an off-the-shelf programmable gate array (Field Programmable Gate Array, FPGA) or other programmable logic. Devices, discrete gate or transistor logic devices, discrete hardware components. Each method, step and logic block diagram disclosed in the embodiments of this specification can be implemented or executed. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of this specification can be directly implemented by a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software module can be located in random access memory, flash memory, read-only memory, programmable read-only memory or electrically erasable programmable memory, registers and other mature storage media in this field. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.
可以理解，本说明书实施方式中的存储器可以是易失性存储器或非易失性存储器，或可包括易失性和非易失性存储器两者。其中，非易失性存储器可以是只读存储器(ROM)、可编程只读存储器(programmable ROM，PROM)、可擦除可编程只读存储器(erasablePROM，EPROM)、电可擦除可编程只读存储器(EEPROM)或闪存。易失性存储器可以是随机存取存储器(RAM)。应注意，本文描述的系统和方法的存储器旨在包括但不限于这些和任意其它适合类型的存储器。It can be understood that the memory in the embodiments of this specification may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable ROM (PROM), an erasable programmable read-only memory (erasablePROM, EPROM), an electrically erasable programmable read-only memory. memory (EEPROM) or flash memory. Volatile memory may be random access memory (RAM). It should be noted that the memory of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.
本领域普通技术人员可以意识到，结合本文中所公开的实施方式描述的各示例的单元及算法步骤，能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本说明书的范围。Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented with electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Skilled artisans may implement the described functions using different methods for each specific application, but such implementations should not be considered beyond the scope of this specification.
所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述的系统、装置和单元的具体工作过程，可以参考前述方法实施方式中的对应过程，在此不再赘述。Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, devices and units described above can be referred to the corresponding processes in the foregoing method implementations, and will not be described again here.
在本说明书所提供的几个实施方式中，应所述理解到，所揭露的系统、装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施方式仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。Among the several implementations provided in this specification, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device implementation described above is only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or can be integrated into another system, or some features can be ignored, or not implemented. On the other hand, the coupling or direct coupling or communication connection between each other shown or discussed may be through some interfaces, and the indirect coupling or communication connection of the devices or units may be in electrical, mechanical or other forms.
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施方式方案的目的。The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or they may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of this embodiment.
另外，在本说明书各个实施方式中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。In addition, each functional unit in each embodiment of this specification may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit.
所述功能如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。基于这样的理解，本说明书的技术方案本质上或者说对现有技术做出贡献的部分或者所述技术方案的部分可以以软件产品的形式体现出来，所述计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本说明书各个实施方式所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(ROM)、随机存取存储器(RAM)、磁碟或者光盘等各种可以存储程序代码的介质。If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution in this specification may be embodied in the form of a software product in essence or the part that contributes to the existing technology or the part of the technical solution, and the computer software product is stored in a storage medium , including several instructions to cause a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of this specification. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk and other various media that can store program codes.
以上所述，仅为本说明书的具体实施方式，但本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本说明书揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本说明书的保护范围之内。因此，本发明的保护范围应所述以权利要求的保护范围为准。The above are only specific embodiments of this specification, but the protection scope of the present invention is not limited thereto. Any person familiar with the technical field can easily think of changes or replacements within the technical scope disclosed in this specification. should be covered by the protection scope of this manual. Therefore, the protection scope of the present invention should be determined by the protection scope of the claims.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310666660.2ACN116394980B (en) | 2023-06-07 | 2023-06-07 | Vehicle control method, automatic driving prompting method and related devices |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310666660.2ACN116394980B (en) | 2023-06-07 | 2023-06-07 | Vehicle control method, automatic driving prompting method and related devices |
| Publication Number | Publication Date |
|---|---|
| CN116394980A CN116394980A (en) | 2023-07-07 |
| CN116394980Btrue CN116394980B (en) | 2023-10-03 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310666660.2AActiveCN116394980B (en) | 2023-06-07 | 2023-06-07 | Vehicle control method, automatic driving prompting method and related devices |
| Country | Link |
|---|---|
| CN (1) | CN116394980B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111123952A (en)* | 2019-12-31 | 2020-05-08 | 华为技术有限公司 | A kind of trajectory planning method and device |
| CN112046502A (en)* | 2019-05-20 | 2020-12-08 | 现代摩比斯株式会社 | Autopilot device and method |
| CN114509065A (en)* | 2022-02-16 | 2022-05-17 | 北京易航远智科技有限公司 | Map construction method, map construction system, vehicle terminal, server side and storage medium |
| CN114771563A (en)* | 2022-04-06 | 2022-07-22 | 扬州大学 | Method for realizing planning control of track of automatic driving vehicle |
| CN115905449A (en)* | 2022-12-30 | 2023-04-04 | 北京易航远智科技有限公司 | Semantic map construction method and automatic driving system with familiar road mode |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6663835B2 (en)* | 2016-10-12 | 2020-03-13 | 本田技研工業株式会社 | Vehicle control device |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112046502A (en)* | 2019-05-20 | 2020-12-08 | 现代摩比斯株式会社 | Autopilot device and method |
| CN111123952A (en)* | 2019-12-31 | 2020-05-08 | 华为技术有限公司 | A kind of trajectory planning method and device |
| CN114509065A (en)* | 2022-02-16 | 2022-05-17 | 北京易航远智科技有限公司 | Map construction method, map construction system, vehicle terminal, server side and storage medium |
| CN114771563A (en)* | 2022-04-06 | 2022-07-22 | 扬州大学 | Method for realizing planning control of track of automatic driving vehicle |
| CN115905449A (en)* | 2022-12-30 | 2023-04-04 | 北京易航远智科技有限公司 | Semantic map construction method and automatic driving system with familiar road mode |
| Publication number | Publication date |
|---|---|
| CN116394980A (en) | 2023-07-07 |
| Publication | Publication Date | Title |
|---|---|---|
| CN116394981B (en) | Vehicle control method, automatic driving prompting method and related devices | |
| CN109583151B (en) | Vehicle trajectory prediction method and device | |
| JP6969962B2 (en) | Map information providing system for vehicle driving support and / or driving control | |
| EP3647733A1 (en) | Automatic annotation of environmental features in a map during navigation of a vehicle | |
| CN117053813A (en) | Navigating a vehicle using an electronic horizon | |
| US20180237019A1 (en) | Autonomous driving assistance system, autonomous driving assistance method, and computer program | |
| DK201970128A1 (en) | AUTOMATIC ANNOTATION OF ENVIRONMENTAL FEATURES IN A MAP DURING NAVIGATION OF A VEHICLE | |
| JP7505448B2 (en) | Driving Support Devices | |
| KR20180009755A (en) | Lane estimation method | |
| JP7405012B2 (en) | Driving support equipment and computer programs | |
| KR102624829B1 (en) | Method, apparatus and computer program for providing route guidance service using location information of vehicle | |
| CN112781600B (en) | Vehicle navigation method, device and storage medium | |
| US12296858B2 (en) | Hybrid log simulated driving | |
| CN116266380A (en) | Environment data reconstruction method, device, system and storage medium | |
| CN118435250A (en) | Stereo auxiliary network for determining the position of objects | |
| JP7501039B2 (en) | Driving assistance device and computer program | |
| JP5790224B2 (en) | Map display system, method and program | |
| JP7662636B2 (en) | Perceptual Error Model | |
| CN116394980B (en) | Vehicle control method, automatic driving prompting method and related devices | |
| CN116821266B (en) | Semantic map construction method and automatic driving system with acquaintance road mode | |
| JP7739909B2 (en) | Driving assistance device and computer program | |
| JP7484317B2 (en) | Driving assistance device and computer program | |
| JP7141480B2 (en) | map generator | |
| CN111637898B (en) | Processing method and device for high-precision navigation electronic map | |
| WO2024162349A1 (en) | Driving assistance device and computer program |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |