技术领域Technical Field
本申请属于智能交通技术领域，具体涉及一种基于数据驱动的混合交通环境网联车辆控制系统及方法。The present application belongs to the field of intelligent transportation technology, and specifically relates to a data-driven mixed traffic environment networked vehicle control system and method.
背景技术Background technique
智能网联车辆技术的快速发展，为实现更优异的车辆控制性能和提高交通系统效率提供了契机。通过车车通信或车辆与基础设施通信，智能网联车辆可以访问超视距信息和边缘/云计算资源，从而做出更复杂的决策，更进一步，多辆智能网联车辆之间可以互相协作，从而实现对整个交通系统的优化。一种典型的智能网联车辆技术是协同自适应巡航控制，它将一系列智能网联车辆组织成队列，并应用协同控制策略来实现更小的跟车间距、更好的燃油经济性和更高效的交通系统。然而，在实际的智能网联车辆技术推广过程中，势必会存在一个长期的过渡阶段，即智能网联车辆与人工驾驶车辆共存的混合交通环境。The rapid development of intelligent connected vehicle technology provides an opportunity to achieve better vehicle control performance and improve the efficiency of the traffic system. Through vehicle-to-vehicle communication or vehicle-to-infrastructure communication, intelligent connected vehicles can access beyond-line-of-sight information and edge/cloud computing resources to make more complex decisions. Furthermore, multiple intelligent connected vehicles can collaborate with each other to optimize the entire traffic system. A typical intelligent connected vehicle technology is cooperative adaptive cruise control, which organizes a series of intelligent connected vehicles into a queue and applies a cooperative control strategy to achieve a smaller following distance, better fuel economy and a more efficient traffic system. However, in the actual promotion of intelligent connected vehicle technology, there will inevitably be a long transition period, that is, a mixed traffic environment where intelligent connected vehicles and manually driven vehicles coexist.
现有混合交通环境网联车辆队列纵向运动控制技术多假设队列系统模型已知，依此进行控制器的设计。然而，由于人工驾驶员的跟车行为是非线性的，并且不同的驾驶员有不同的驾驶风格，在实际混合交通环境中，车辆队列系统模型是无法提前获得的。设计针对混合交通环境下未知车辆队列模型的纵向运动控制技术，对于提升过渡阶段智能网联汽车对交通系统的积极作用具有重要意义。Existing longitudinal motion control technologies for connected vehicle platoons in mixed traffic environments mostly assume that the platoon system model is known, and the controller is designed based on this. However, since the following behavior of human drivers is nonlinear and different drivers have different driving styles, the vehicle platoon system model cannot be obtained in advance in actual mixed traffic environments. Designing longitudinal motion control technology for unknown vehicle platoon models in mixed traffic environments is of great significance for enhancing the positive role of intelligent connected vehicles in the transition stage on the transportation system.
现有技术一公开了一种考虑后车信息的混合车辆队列控制方法及系统(CN112907937B)，其揭示混合车辆队列由沿纵向排列行驶的n+1辆车组成，按照从前到后的顺序依次采用序号0到n进行编号，0号车为可自主驾驶的领航车，1～n号车为跟驰车，k号跟驰车为智能网联车辆，其余的所述跟驰车为人工驾驶车辆；混合车辆队列控制方法包括：人工驾驶车辆跟随前车行驶；利用预设公式控制所述k号跟驰车。该发明能够保证混合车辆队列系统渐近稳定性与首尾队列稳定性，实现混合车辆队列跟驰目标。但是，该技术方案假设混合车辆队列的系统模型是已知的，并以此设计控制器。但是在实际场景中混合车辆队列的模型难以直接获得。Prior art 1 discloses a method and system for controlling a mixed vehicle queue considering the information of the following vehicle (CN112907937B), which discloses that the mixed vehicle queue is composed of n+1 vehicles arranged in a longitudinal direction, and are numbered from 0 to n in order from front to back, vehicle No. 0 is a pilot vehicle that can be driven autonomously, vehicles No. 1 to No. n are following vehicles, No. k is an intelligent networked vehicle, and the remaining following vehicles are manually driven vehicles; the mixed vehicle queue control method includes: a manually driven vehicle follows the leading vehicle; and the No. k following vehicle is controlled by a preset formula. The invention can ensure the asymptotic stability of the mixed vehicle queue system and the stability of the head and tail queues, and achieve the mixed vehicle queue following target. However, the technical solution assumes that the system model of the mixed vehicle queue is known, and designs the controller based on it. However, in actual scenarios, the model of the mixed vehicle queue is difficult to obtain directly.
现有技术二公开了一种混合车辆队列分布式模型预测控制方法(CN113791615A，其揭示首先将混合车辆队列划分为多个相互关联的局部混合车辆队列；然后基于网联自动驾驶车辆和人类驾驶车辆的跟驰模型，建立各个局部混合队列的模型；其次，基于局部混合队列模型，根据模型预测控制算法，建立混合车辆队列整体的控制问题；最后，结合交替方向乘子法，为各个局部混合队列构建分布式模型预测控制器，实现混合车辆队列整体的编队控制。但是，该技术方案假设混合车辆队列的系统模型是已知的，并以此设计控制器。但是在实际场景中混合车辆队列的模型难以直接获得。Prior art 2 discloses a distributed model predictive control method for a mixed vehicle queue (CN113791615A, which discloses that firstly, a mixed vehicle queue is divided into a plurality of interrelated local mixed vehicle queues; then, based on the following model of the networked autonomous driving vehicle and the human-driven vehicle, a model of each local mixed queue is established; secondly, based on the local mixed queue model, according to the model predictive control algorithm, a control problem of the whole mixed vehicle queue is established; finally, in combination with the alternating direction multiplier method, a distributed model predictive controller is constructed for each local mixed queue to realize the formation control of the whole mixed vehicle queue. However, the technical solution assumes that the system model of the mixed vehicle queue is known, and designs the controller based on it. However, in actual scenarios, the model of the mixed vehicle queue is difficult to obtain directly.
因此，迫切需要开发出一种基于数据驱动的混合交通环境网联车辆控制系统及方法以解决上述技术问题。Therefore, there is an urgent need to develop a data-driven mixed traffic environment connected vehicle control system and method to solve the above technical problems.
发明内容Summary of the invention
本申请实施例的目的是提供一种基于数据驱动的混合交通环境网联车辆控制系统及方法，其设计一种数据驱动的控制算法，利用车辆队列行驶过程中收集到的状态数据，可以直接得到混合车辆队列系统的控制器并辨识出混合车辆队列系统模型，辨识后的模型可以用于设计更复杂的多目标优化的控制器，从而可以有效解决背景技术中涉及的至少一个技术问题。The purpose of the embodiments of the present application is to provide a data-driven mixed traffic environment networked vehicle control system and method, which designs a data-driven control algorithm, and uses the state data collected during the vehicle queue driving process to directly obtain the controller of the mixed vehicle queue system and identify the mixed vehicle queue system model. The identified model can be used to design a more complex multi-objective optimization controller, thereby effectively solving at least one technical problem involved in the background technology.
为了解决上述技术问题，本申请是这样实现的：In order to solve the above technical problems, this application is implemented as follows:
本申请实施例提供了一种基于数据驱动的混合交通环境网联车辆控制方法，包括：The embodiment of the present application provides a data-driven connected vehicle control method in a mixed traffic environment, including:
步骤S1：对混合车辆队列进行建模，包括：Step S1: Modeling a mixed vehicle platoon, including:
步骤S11：对在平直道路上行驶的随机组成的包含n+1辆车的混合车辆队列，其包含1辆作为领航车的0号车，m辆智能网联车辆，n-m辆人工驾驶车辆，其中，队列中第i辆车的位置、速度、加速度分别用pi、vi和ai表示，i∈1，2，...，n；si＝pi-pi-1表示第i辆车与前车间的跟车间距；Step S11: for a mixed vehicle platoon consisting of n+1 vehicles randomly formed and traveling on a straight road, which includes a No. 0 vehicle as a pilot vehicle, m intelligent network-connected vehicles, and nm manually driven vehicles, the position, speed, and acceleration of the i-th vehicle in the platoon are represented bypi , vi, andai, respectively, i∈1,2,...,n;si =pi -pi-1 represents the following distance between the i-th vehicle and the vehicle in front;
步骤S12：设集合N＝1，2，...，n，表示除领航车外，队列中其他车辆下标的集合，M＝{i1，i2，...，im}为队列中智能网联车辆下标的集合，那么，对于人工驾驶车辆，采用最优速度模型描述其跟驰行为：Step S12: Let N = 1, 2, ..., n, which represents the set of indices of other vehicles in the queue except the pilot vehicle, and M = {i1 , i2 , ..., im }, which represents the set of indices of intelligent networked vehicles in the queue. Then, for manually driven vehicles, the optimal speed model is used to describe their following behavior:
其中，α，β＞0为驾驶员的反应参数，α为驾驶员对跟车间距的反应增益，β为驾驶员对自车与前车速度差的反应增益，vdes(s)是与跟车间距相关的驾驶员期望速度，一般表述为连续的分段函数：Among them, α, β>0 are the driver's reaction parameters, α is the driver's reaction gain to the following distance, β is the driver's reaction gain to the speed difference between the vehicle and the front vehicle, and vdes (s) is the driver's expected speed related to the following distance, which is generally expressed as a continuous piecewise function:
步骤S13：定义s*、v*为混合车辆队列中每辆车要达到的期望间距和期望速度，二者之间的关系为v*＝vdes(s*)；Step S13: define s* and v* as the expected spacing and expected speed to be achieved by each vehicle in the mixed vehicle queue, and the relationship between the two is v* = vdes (s* );
步骤S14：对智能网联车辆采用二阶动力学模型进行建模：Step S14: Modeling the intelligent connected vehicle using a second-order dynamics model:
其中，i∈N，表示第i辆车在t时刻的跟车间距误差和速度误差；对于领航车，定义∈(t)＝v0(t)-v*；ui(t)为第i辆智能网联车辆的加速度；in, i∈N represents the following distance error and speed error of the i-th vehicle at time t; for the pilot vehicle, define ∈(t)=v0 (t)-v* ;ui (t) is the acceleration of the i-th intelligent connected vehicle;
得到混合车辆队列模型的状态空间方程表达式：The state space equation expression of the mixed vehicle platoon model is obtained:
其中，in,
α1＝αvdes(s*)，α2＝α+β，α3＝βα1 = αvdes (s* ), α2 = α + β, α3 = β
其中，表示在一个n×1的向量中，第i个元素为1，其他元素全为0；为状态量；in, Indicates that in an n×1 vector, the i-th element is 1 and all other elements are 0; is the state quantity;
利用前向欧拉法将状态空间方程离散化，采样时间间隔为Ts：The state space equation is discretized using the forward Euler method, with a sampling time interval of Ts :
x(t+Ts)＝Adx(t)+Bdu(t)+Hd∈(t)；x(t+Ts )=Adx (t)+Bdu (t)+Hd∈ (t);
步骤S2：在混合车辆队列运行的初始阶段，智能网联车辆运行在自适应巡航控制模式，并在控制量中加入探索噪声e，即ui(t)＝uACC+e，i∈M.收集采样l≥4n2+4mn个时间Ts的数据，构造δxx、Ixx、Ixu、X、X′、γ矩阵：Step S2: In the initial stage of the mixed vehicle platoon operation, the intelligent connected vehicles operate in the adaptive cruise control mode, and add the exploration noise e to the control quantity, that is,ui (t) =uACC + e, i∈M. Collect and sample data ofl≥4n2 +4mn timeTs , and constructδxx ,Ixx ,Ixu , X, X′, γ matrices:
其中，为克罗内可积算子，vec(X)表示矩阵X向量化；/>为单in, is the Krone integrable operator, vec(X) represents the vectorization of matrix X;/> For single
位矩阵，为对阵正定矩阵；Kk为状态反馈矩阵，下标k∈{0，1，2，...，kmax}表示迭代次数，初始化K0＝0，k＝0；bit matrix, is a positive definite matrix for the pair; Kk is a state feedback matrix, subscript k∈{0, 1, 2, ...,kmax } represents the number of iterations, initialized with K0 = 0, k = 0;
步骤S3：求解对称正定矩阵Pk和状态反馈矩阵Kk+1，实现策略评估和策略提升：Step S3: Solve the symmetric positive definite matrix Pk and the state feedback matrix Kk+1 to achieve strategy evaluation and strategy improvement:
其中，k表示迭代次数；Where k represents the number of iterations;
步骤S4：令k＝k+1，并重复步骤S3，直至|Pk+1-Pk|＜τ或k＞kmax.τ为预设的迭代容许阈值；Step S4: let k=k+1, and repeat step S3 until |Pk+1 -Pk |＜τ or k＞kmax .τ is a preset iteration tolerance threshold;
步骤S5：更新控制器，将u＝-Kkx作为近似最优控制律用于混合车辆队列；Step S5: Update the controller and use u=-Kk x as the approximate optimal control law for the mixed vehicle platoon;
步骤S6：利用参数辨识或非参数化建模的方法构建混合车辆队列系统模型，通过参数辨识估计Ad、Bd、Hd矩阵或者通过非参数化的形式，对混合队列系统建模；Step S6: constructing a mixed vehicle platoon system model by using parameter identification or non-parametric modeling methods, estimating Ad , Bd , Hd matrices by parameter identification or modeling the mixed platoon system by non-parametric form;
步骤S7：根据构建的混合车辆队列模型设计控制器。Step S7: Design a controller according to the constructed mixed vehicle platoon model.
可选的，在步骤S6中，Ad、Bd、Hd矩阵由以下公式计算：Optionally, in step S6, the Ad , Bd , and Hd matrices are calculated using the following formula:
X＝U∑VTX＝U∑VT
其中，为Ad、Bd矩阵的近似值，/>∑除了主对角线上的元素以外全为0，主对角线上的每个元素都称为奇异值，U、V满足UTU＝I、VTV＝I，I为单位矩阵；Hd矩阵由α3和0构成，α3可从/>中获得。in, is the approximate value of the Ad and Bd matrices,/> ∑ All elements except those on the main diagonal are 0. Each element on the main diagonal is called a singular value. U and V satisfy UT U = I and VT V = I, where I is the unit matrix. The Hd matrix is composed of α3 and 0. α3 can be obtained from/> Obtained in.
可选的，步骤S7具体包括：Optionally, step S7 specifically includes:
利用模型预测控制的方法求解下述优化问题：The model predictive control method is used to solve the following optimization problem:
subject to：subject to：
x(k+1|t)＝Adx(k|t)+Bdu(k|t)x(k+1|t)＝Ad x(k|t)+Bd u(k|t)
amin≤u(k|t)≤amaxamin ≤u(k|t)≤amax
其中，J为代价函数，Np为预测时域，x(k|t)为在当前时刻t预测t+k时刻的状态，u(k|t)为在当前时刻t预测t+k时刻的控制输入，amin、amax为控制输入约束的上下界，为在当前时刻t预测t+k时刻的智能网联车辆的跟车间距误差，/>为保证安全性所设计的跟车间距误差约束上下界。Where J is the cost function,Np is the prediction time domain, x(k|t) is the state predicted at time t+k from the current time t, u(k|t) is the control input predicted at time t+k from the current time t,amin andamax are the upper and lower bounds of the control input constraints, is the following distance error of the intelligent connected vehicle predicted at time t+k at the current time t,/> The upper and lower bounds of the following vehicle distance error constraints are designed to ensure safety.
本申请实施例还提供了一种用于实现所述的方法的基于数据驱动的混合交通环境网联车辆控制系统，包括安装于混合车辆队列中所有车辆上的车间通信单元、数据存储单元、计算单元以及执行单元，其中，The embodiment of the present application also provides a data-driven mixed traffic environment networked vehicle control system for implementing the method, comprising a vehicle communication unit, a data storage unit, a computing unit, and an execution unit installed on all vehicles in a mixed vehicle queue, wherein:
车间通信单元，用于向外传递自车或接收混合车辆队列中其他车辆的位置、速度和加速度数据；A vehicle-to-vehicle communication unit, used to transmit the position, speed and acceleration data of the vehicle itself or receive the position, speed and acceleration data of other vehicles in the mixed vehicle queue;
数据存储单元，用于储存所述的位置、速度和加速度数据，并供计算单元读取和写入新的数据；A data storage unit, used to store the position, velocity and acceleration data, and for the calculation unit to read and write new data;
计算单元，用于处理收集到的数据向执行单元发送控制指令以执行所述的方法；A computing unit, configured to process the collected data and send a control instruction to an execution unit to execute the method;
执行单元，通过调节自车驱动或制动力实现执行单元下发的控制指令。The execution unit implements the control instructions issued by the execution unit by adjusting the driving or braking force of the vehicle.
本申请具有如下有益效果：能有效解决实际混合交通场景中车辆队列模型未知，传统的基于模型进行控制器设计的方法无法使用的问题。在车辆队列运行的初始阶段收集数据通过所提技术方案就能得到保证车辆队列内稳定性的控制器和混合车辆队列的近似模型。更进一步，依据辨识出的混合队列模型可以设计更复杂的控制器用于保证智能网联汽车的跟车安全性和舒适性。The present application has the following beneficial effects: it can effectively solve the problem that the vehicle queue model in actual mixed traffic scenarios is unknown and the traditional model-based controller design method cannot be used. By collecting data in the initial stage of vehicle queue operation, the controller that ensures the stability of the vehicle queue and the approximate model of the mixed vehicle queue can be obtained through the proposed technical solution. Furthermore, based on the identified mixed queue model, a more complex controller can be designed to ensure the safety and comfort of the following vehicle of the intelligent connected vehicle.
附图说明BRIEF DESCRIPTION OF THE DRAWINGS
图1是本申请实施例提供的基于数据驱动的混合交通环境网联车辆控制方法的流程图；FIG1 is a flow chart of a method for controlling a connected vehicle in a mixed traffic environment based on data-driven control according to an embodiment of the present application;
图2是本申请实施例提供的混合车辆队列示意图。FIG. 2 is a schematic diagram of a mixed vehicle queue provided in an embodiment of the present application.
具体实施方式Detailed ways
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，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，都属于本申请保护的范围。The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
本申请的说明书和权利要求书中的术语“第一”、“第二”等是用于区别类似的对象，而不用于描述特定的顺序或先后次序。应该理解这样使用的数据在适当情况下可以互换，以便本申请的实施例能够以除了在这里图示或描述的那些以外的顺序实施，且“第一”、“第二”等所区分的对象通常为一类，并不限定对象的个数，例如第一对象可以是一个，也可以是多个。此外，说明书以及权利要求中“和/或”表示所连接对象的至少其中之一，字符“/”，一般表示前后关联对象是一种“或”的关系。The terms "first", "second", etc. in the specification and claims of this application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of one type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and/or" in the specification and claims represents at least one of the connected objects, and the character "/" generally indicates that the objects associated with each other are in an "or" relationship.
下面结合附图，通过具体的实施例及其应用场景对本申请实施例提供的基于数据驱动的混合交通环境网联车辆控制方法进行详细地说明。The data-driven mixed traffic environment connected vehicle control method provided by the embodiment of the present application is described in detail below with reference to the accompanying drawings through specific embodiments and their application scenarios.
请参见图1和2所示，是本申请实施例提供的一种基于数据驱动的混合交通环境网联车辆控制方法，包括：Please refer to Figures 1 and 2, which are a data-driven mixed traffic environment connected vehicle control method provided by an embodiment of the present application, including:
步骤S1：对混合车辆队列进行建模，包括：Step S1: Modeling a mixed vehicle platoon, including:
步骤S11：对在平直道路上行驶的随机组成的包含n+1辆车的混合车辆队列，其包含1辆作为领航车的0号车，m辆智能网联车辆，n-m辆人工驾驶车辆，其中，队列中第i辆车的位置、速度、加速度分别用pi、vi和ai表示，i∈1，2，...，n；si＝pi-pi-1表示第i辆车与前车间的跟车间距；Step S11: for a mixed vehicle platoon consisting of n+1 vehicles randomly formed and traveling on a straight road, which includes a No. 0 vehicle as a pilot vehicle, m intelligent network-connected vehicles, and nm manually driven vehicles, the position, speed, and acceleration of the i-th vehicle in the platoon are represented bypi , vi, andai, respectively, i∈1,2,...,n;si =pi -pi-1 represents the following distance between the i-th vehicle and the vehicle in front;
步骤S12：设集合N＝1，2，...，n，表示除领航车外，队列中其他车辆下标的集合，M＝{i1，i2，...，im)为队列中智能网联车辆下标的集合，那么，对于人工驾驶车辆，采用最优速度模型描述其跟驰行为：Step S12: Let N = 1, 2, ..., n, which represents the set of indices of other vehicles in the queue except the pilot vehicle, and M = {i1 , i2 , ..., im ) is the set of indices of intelligent networked vehicles in the queue. Then, for manually driven vehicles, the optimal speed model is used to describe their following behavior:
其中，α，β＞0为驾驶员的反应参数，α为驾驶员对跟车间距的反应增益，β为驾驶员对自车与前车速度差的反应增益，vdes(s)是与跟车间距相关的驾驶员期望速度，一般表述为连续的分段函数：Among them, α, β>0 are the driver's reaction parameters, α is the driver's reaction gain to the following distance, β is the driver's reaction gain to the speed difference between the vehicle and the front vehicle, and vdes (s) is the driver's expected speed related to the following distance, which is generally expressed as a continuous piecewise function:
该式说明当车辆与前车的跟车间距s小于等于st时，车辆因为安全原因而停车；而当车辆与前车距离s大于等于sgo时，车辆保持预设的最大车速vmax辆与前车的跟车间距在sst和sgo之间时，驾驶员的期望速度随着跟车间距单调增加。This formula shows that when the following distance s between the vehicle and the vehicle in front is less than or equal to st , the vehicle stops for safety reasons; when the distance s between the vehicle and the vehicle in front is greater than or equal to sgo , the vehicle maintains the preset maximum speed vmax. When the following distance between the vehicle and the vehicle in front is between sst and sgo , the driver's expected speed increases monotonically with the following distance.
步骤S13：定义s*、v*为混合车辆队列中每辆车要达到的期望间距和期望速度，二者之间的关系为v*＝vdes(s*)；Step S13: define s* and v* as the expected spacing and expected speed to be achieved by each vehicle in the mixed vehicle queue, and the relationship between the two is v* = vdes (s* );
步骤S14：对智能网联车辆采用二阶动力学模型进行建模：Step S14: Modeling the intelligent connected vehicle using a second-order dynamics model:
其中，i∈N，表示第i辆车在t时刻的跟车间距误差和速度误差；对于领航车，定义∈(t)＝v0(t)-v*；ui(t)为第i辆智能网联车辆的加速度；in, i∈N represents the following distance error and speed error of the i-th vehicle at time t; for the pilot vehicle, define ∈(t)=v0 (t)-v* ;ui (t) is the acceleration of the i-th intelligent connected vehicle;
得到混合车辆队列模型的状态空间方程表达式：The state space equation expression of the mixed vehicle platoon model is obtained:
其中，in,
α1＝αvdes(s*)，α2＝α+β，α3＝βα1 = αvdes (s* ), α2 = α + β, α3 = β
其中，表示在一个n×1的向量中，第i个元素为1，其他元素全为0；为状态量；in, Indicates that in an n×1 vector, the i-th element is 1 and all other elements are 0; is the state quantity;
利用前向欧拉法将状态空间方程离散化，采样时间间隔为Ts：The state space equation is discretized using the forward Euler method, with a sampling time interval of Ts :
x(t+Ts)＝Adx(t)+Bdu(t)+Hd∈(t)x(t+Ts )＝Ad x(t)+Bd u(t)+Hd ∈(t)
需要注意的是，在混合车辆队列运行的初始阶段，A、B、H矩阵均是未知的，因此在运行的初始阶段取v*＝v0(t)。It should be noted that, in the initial stage of the operation of the mixed vehicle platoon, the A, B, and H matrices are all unknown, so v* = v0 (t) is taken in the initial stage of the operation.
步骤S2：在混合车辆队列运行的初始阶段，智能网联车辆运行在自适应巡航控制模式，并在控制量中加入探索噪声，即ui(t)＝uACC+e，i∈M.e为探索噪声，其形式有随机噪声、指数递减噪声、正\余弦函数噪声等，收集采样l≥4n2+4mn个时间Ts的数据，构造δxx、Ixx、Ixu、X、X′、γ矩阵：Step S2: In the initial stage of the mixed vehicle platoon operation, the intelligent connected vehicles operate in the adaptive cruise control mode, and add exploration noise to the control quantity, that is,ui (t) =uACC + e, i∈Me is the exploration noise, which has the form of random noise, exponentially decreasing noise, sine\cosine function noise, etc., collect and sample data ofl≥4n2 +4mn timeTs , and constructδxx ,Ixx ,Ixu , X, X′, γ matrices:
其中，为克罗内可积算子，vec(X)表示矩阵X向量化；/>为单位矩阵，/>为对阵正定矩阵；Kk为状态反馈矩阵，下标k∈{0，1，2，...，kmax}表示迭代次数，初始化K0＝0，k＝0；in, is the Krone integrable operator, vec(X) represents the vectorization of matrix X;/> is the identity matrix, /> is a positive definite matrix for the pair; Kk is a state feedback matrix, subscript k∈{0, 1, 2, ...,kmax } represents the number of iterations, initialized with K0 = 0, k = 0;
步骤S3：求解对称正定矩阵Pk和状态反馈矩阵Kk+1，实现策略评估和策略提升：Step S3: Solve the symmetric positive definite matrix Pk and the state feedback matrix Kk+1 to achieve strategy evaluation and strategy improvement:
其中，k表示迭代次数；Where k represents the number of iterations;
步骤S4：令k＝k+1，并重复步骤S3，直至|Pk+1-Pk|＜ε或k＞kmax.ε为预设的迭代容许阈值；Step S4: let k=k+1, and repeat step S3 until |Pk+1 -Pk |＜ε or k＞kmax .ε is a preset iteration tolerance threshold;
步骤S5：更新控制器，将u＝-Kkx作为近似最优控制律用于混合车辆队列；Step S5: Update the controller and use u=-Kk x as the approximate optimal control law for the mixed vehicle platoon;
步骤S6：利用参数辨识或非参数化建模的方法构建混合车辆队列系统模型，通过参数辨识估计Ad、Bd、Hd矩阵或者通过非参数化的形式，对混合队列系统建模；Step S6: constructing a mixed vehicle platoon system model by using parameter identification or non-parametric modeling methods, estimating Ad , Bd , Hd matrices by parameter identification or modeling the mixed platoon system by non-parametric form;
步骤S7：根据构建的混合车辆队列模型设计控制器。Step S7: Design a controller according to the constructed mixed vehicle platoon model.
需要进一步说明的是，在步骤S6中，Ad、Bd、Hd矩阵由以下公式计算：It should be further explained that, in step S6, the Ad , Bd , and Hd matrices are calculated by the following formula:
X＝U∑VTX＝U∑VT
其中，为Ad、Bd矩阵的近似值，/>∑除了主对角线上的元素以外全为0，主对角线上的每个元素都称为奇异值，U、V满足UTU＝I、VTV＝I，I为单位矩阵；Hd矩阵由α3和0构成，α3可从/>中获得。in, is the approximate value of the Ad and Bd matrices,/> ∑ All elements except those on the main diagonal are 0. Each element on the main diagonal is called a singular value. U and V satisfy UT U = I and VT V = I, where I is the unit matrix. The Hd matrix is composed of α3 and 0. α3 can be obtained from/> Obtained in.
步骤S7具体包括：Step S7 specifically includes:
利用模型预测控制的方法求解下述优化问题：The model predictive control method is used to solve the following optimization problem:
subject to：subject to：
x(k+1|t)＝Adx(k|t)+Bau(k|t)x(k+1|t)＝Ad x(k|t)+Ba u(k|t)
amin≤u(k|t)≤amaxamin ≤u(k|t)≤amax
其中，J为代价函数，Np为预测时域，x(k|t)为在当前时刻t预测t+k时刻的状态，u(k|t)为在当前时刻t预测t+k时刻的控制输入，amin、amax为控制输入约束的上下界，为在当前时刻t预测t+k时刻的智能网联车辆的跟车间距误差，/>为保证安全性所设计的跟车间距误差约束上下界。Where J is the cost function,Np is the prediction time domain, x(k|t) is the state predicted at time t+k from the current time t, u(k|t) is the control input predicted at time t+k from the current time t,amin andamax are the upper and lower bounds of the control input constraints, is the following distance error of the intelligent connected vehicle predicted at time t+k at the current time t,/> The upper and lower bounds of the following vehicle distance error constraints are designed to ensure safety.
本申请实施例还提供了一种用于实现所述的方法的基于数据驱动的混合交通环境网联车辆控制系统，包括安装于混合车辆队列中所有车辆上的车间通信单元、数据存储单元、计算单元以及执行单元，其中，The embodiment of the present application also provides a data-driven mixed traffic environment networked vehicle control system for implementing the method, comprising a vehicle communication unit, a data storage unit, a computing unit, and an execution unit installed on all vehicles in a mixed vehicle queue, wherein:
车间通信单元，用于向外传递自车或接收混合车辆队列中其他车辆的位置、速度和加速度数据；A vehicle-to-vehicle communication unit, used to transmit the position, speed and acceleration data of the vehicle itself or receive the position, speed and acceleration data of other vehicles in the mixed vehicle platoon;
数据存储单元，用于储存所述的位置、速度和加速度数据，并供计算单元读取和写入新的数据；A data storage unit, used to store the position, velocity and acceleration data, and for the calculation unit to read and write new data;
计算单元，用于处理收集到的数据向执行单元发送控制指令以执行所述的方法；A computing unit, configured to process the collected data and send a control instruction to an execution unit to execute the method;
执行单元，通过调节自车驱动或制动力实现执行单元下发的控制指令。The execution unit implements the control instructions issued by the execution unit by adjusting the driving or braking force of the vehicle.
本申请具有如下有益效果：能有效解决实际混合交通场景中车辆队列模型未知，传统的基于模型进行控制器设计的方法无法使用的问题。在车辆队列运行的初始阶段收集数据通过所提技术方案就能得到保证车辆队列内稳定性的控制器和混合车辆队列的近似模型。更进一步，依据辨识出的混合队列模型可以设计更复杂的控制器用于保证智能网联汽车的跟车安全性和舒适性。The present application has the following beneficial effects: it can effectively solve the problem that the vehicle queue model in actual mixed traffic scenarios is unknown and the traditional model-based controller design method cannot be used. By collecting data in the initial stage of vehicle queue operation, the controller that ensures the stability of the vehicle queue and the approximate model of the mixed vehicle queue can be obtained through the proposed technical solution. Furthermore, based on the identified mixed queue model, a more complex controller can be designed to ensure the safety and comfort of the following vehicle of the intelligent connected vehicle.
上面结合附图对本申请的实施例进行了描述，但是本申请并不局限于上述的具体实施方式，上述的具体实施方式仅仅是示意性的，而不是限制性的，本领域的普通技术人员在本申请的启示下，在不脱离本申请宗旨和权利要求所保护的范围情况下，还可做出很多形式，均属于本申请的保护之内。The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310531628.3ACN116758722B (en) | 2023-05-12 | 2023-05-12 | Data-driven hybrid traffic environment network-connected vehicle control system and method |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310531628.3ACN116758722B (en) | 2023-05-12 | 2023-05-12 | Data-driven hybrid traffic environment network-connected vehicle control system and method |
| Publication Number | Publication Date |
|---|---|
| CN116758722A CN116758722A (en) | 2023-09-15 |
| CN116758722Btrue CN116758722B (en) | 2024-05-31 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310531628.3AActiveCN116758722B (en) | 2023-05-12 | 2023-05-12 | Data-driven hybrid traffic environment network-connected vehicle control system and method |
| Country | Link |
|---|---|
| CN (1) | CN116758722B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117373235B (en)* | 2023-09-28 | 2024-12-10 | 合肥工业大学 | A method for exploring the equilibrium strategy of interactive game between pedestrians and autonomous vehicles |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108216190A (en)* | 2016-12-19 | 2018-06-29 | 大陆汽车系统公司 | For the predictive power assembly restriction strategy of autonomous/automatic driving vehicle |
| CN111445015A (en)* | 2020-06-15 | 2020-07-24 | 之江实验室 | A method for location estimation of non-connected vehicles in an intelligent networked environment |
| CN111439260A (en)* | 2020-04-27 | 2020-07-24 | 吉林大学 | Cruise driving optimization control system for network-connected commercial diesel vehicles for individual needs |
| CN111586130A (en)* | 2020-04-29 | 2020-08-25 | 清华大学 | Performance Evaluation Method and Device for Longitudinal Control Algorithm in Mixed Traffic Scenario |
| CN113674525A (en)* | 2021-07-30 | 2021-11-19 | 长安大学 | Prediction method of vehicle queuing length at signalized intersection based on sparse data |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020014128A1 (en)* | 2018-07-10 | 2020-01-16 | Cavh Llc | Vehicle on-board unit for connected and automated vehicle systems |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108216190A (en)* | 2016-12-19 | 2018-06-29 | 大陆汽车系统公司 | For the predictive power assembly restriction strategy of autonomous/automatic driving vehicle |
| CN111439260A (en)* | 2020-04-27 | 2020-07-24 | 吉林大学 | Cruise driving optimization control system for network-connected commercial diesel vehicles for individual needs |
| CN111586130A (en)* | 2020-04-29 | 2020-08-25 | 清华大学 | Performance Evaluation Method and Device for Longitudinal Control Algorithm in Mixed Traffic Scenario |
| CN111445015A (en)* | 2020-06-15 | 2020-07-24 | 之江实验室 | A method for location estimation of non-connected vehicles in an intelligent networked environment |
| CN113674525A (en)* | 2021-07-30 | 2021-11-19 | 长安大学 | Prediction method of vehicle queuing length at signalized intersection based on sparse data |
| Title |
|---|
| Data-Driven Predictive Control for Connected and Autonomous Vehicles in Mixed Traffic;Jiawei Wang等;《2022 American Control Conference (ACC)》;20220905;全文* |
| 智能网联环境下基于混合深度学习的交通流预测模型;陆文琦;芮一康;冉斌;谷远利;;交通运输系统工程与信息;20200615(第03期);全文* |
| 电动汽车自适应巡航控制系统建模与仿真;闫丹彤;何智成;陈东;谭纯;;计算机仿真;20160115(第01期);全文* |
| Publication number | Publication date |
|---|---|
| CN116758722A (en) | 2023-09-15 |
| Publication | Publication Date | Title |
|---|---|---|
| You et al. | Advanced planning for autonomous vehicles using reinforcement learning and deep inverse reinforcement learning | |
| KR102461831B1 (en) | System and Method for Improving of Advanced Deep Reinforcement Learning Based Traffic in Non signalalized Intersections for the Multiple Self driving Vehicles | |
| CN111483468B (en) | A lane-changing decision-making method and system for unmanned vehicles based on adversarial imitation learning | |
| CN112445229B (en) | Single-lane multi-queue hierarchical control method for piloting motorcade cooperation | |
| CN111768616B (en) | Vehicle fleet consistency control method based on vehicle-road cooperation in mixed traffic scene | |
| CN111332290B (en) | A vehicle formation method and system based on feedforward-feedback control | |
| CN106502098B (en) | A kind of optimal speed closed loop quick predict control method and device based on car networking | |
| JP2019001450A (en) | Real-time vehicle state trajectory prediction for vehicle energy management and autonomous drive | |
| CN113791615A (en) | A Distributed Model Predictive Control Method for Hybrid Vehicle Queue | |
| CN116758722B (en) | Data-driven hybrid traffic environment network-connected vehicle control system and method | |
| CN112907937B (en) | Hybrid vehicle queue control method and system considering rear vehicle information | |
| CN116382297A (en) | Constrained Hybrid Vehicle Formation Control Method Based on Deep Reinforcement Learning Strategy | |
| CN114253274A (en) | Data-driven-based online hybrid vehicle formation rolling optimization control method | |
| CN116088530B (en) | Reinforcement learning autonomous driving platoon control method based on model predictive control guidance | |
| CN114519433A (en) | Multi-agent reinforcement learning and strategy execution method and computer equipment | |
| CN113140104B (en) | Vehicle queue tracking control method and device and computer readable storage medium | |
| CN115103313A (en) | Method and system for collaborative processing of smart highway tasks based on location prediction | |
| CN117523813A (en) | A collaborative predictive control method for vehicle queues in curve scenarios | |
| CN119512101A (en) | A multi-vehicle cooperative control method and system considering time-varying directed communication topology | |
| CN118612743B (en) | Vehicle edge computing network client scheduling method and system based on reputation model | |
| CN110456790B (en) | Optimal control method of intelligent networked electric vehicle platoon based on adaptive weight | |
| CN117724493A (en) | Mixed cooperative driving control method and system considering communication time delay | |
| CN115731690B (en) | A decision-making method for unmanned bus clusters based on graph neural network reinforcement learning | |
| CN117289719A (en) | Training method and device for unmanned aerial vehicle control model and unmanned aerial vehicle formation control method | |
| CN115497281B (en) | Robust model predictive control method for multiple queues of pilot vehicles in mixed traffic scenarios |
| 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 | ||
| OL01 | Intention to license declared | ||
| OL01 | Intention to license declared |