技术领域technical field
本发明涉及机器人控制领域，尤其涉及一种基于啮齿类动物模型的同步定位与地图构建方法及装置。The invention relates to the field of robot control, in particular to a method and device for synchronous positioning and map construction based on a rodent model.
背景技术Background technique
同步定位与地图构建是移动机器人目前面临的重大难题。因为移动机器人实质上就是移动的传感器平台，传感器虽然类型和能力各有不同，但是广泛存在里程计漂移和不同的噪声等问题。后来经过学者们的不断探究，仿生机器人(采用仿生技术控制的机器人)逐渐凸显出良好的应用前景，表现出完美的生物合理性和对自然环境的高度适应性。其中，仿生机器人大多使用啮齿类动物模型进行仿生技术实现。Simultaneous localization and map construction are major challenges for mobile robots. Because the mobile robot is essentially a mobile sensor platform, although the types and capabilities of the sensors are different, there are widespread problems such as odometry drift and different noises. Later, after continuous exploration by scholars, bionic robots (robots controlled by bionic technology) gradually showed good application prospects, showing perfect biological rationality and high adaptability to the natural environment. Among them, most bionic robots use rodent animal models for bionic technology.
其中，啮齿类动物模型，将视觉里程计信息和视觉场景图像信息集成到位姿感知细胞模型中，从而使得移动机器人具备一定的更新预测能力，并建立起时间、空间位置、行为等信息的经历图。目前，啮齿类动物模型已经广泛用于机器人的定位导航工作中，解决了众多同步定位与地图构建(Simultaneous Localization and Mapping，MAP)难以解决的问题,但是啮齿类动物模型获取的视觉场景图像信息以及视觉里程计信息均存在一定程度的误差，虽然有学者针对视觉里程计的误差引入FAB-MAP(fast appearance basedmapping)，这种基于历史模型的闭环检测算法，通过实时关键帧的匹配，可以提高系统的稳定性，但是定位的精度并不稳定，且鲁棒性不强。所以，单独的啮齿类动物模型在定位精度及鲁棒性方面有待进一步改善。Among them, the rodent model integrates visual odometer information and visual scene image information into the pose perception cell model, so that the mobile robot has a certain ability to update and predict, and establishes an experience map of information such as time, spatial location, and behavior. . At present, rodent models have been widely used in the positioning and navigation of robots, solving many problems that are difficult to solve in Simultaneous Localization and Mapping (MAP), but the visual scene image information obtained by rodent models and There is a certain degree of error in the visual odometry information. Although some scholars have introduced FAB-MAP (fast appearance based mapping) for the error of the visual odometer, this closed-loop detection algorithm based on the historical model can improve the system by matching key frames in real time. stability, but the positioning accuracy is not stable, and the robustness is not strong. Therefore, the single rodent model needs to be further improved in terms of positioning accuracy and robustness.
发明内容Contents of the invention
本发明要解决的技术问题是提供一种基于啮齿类动物模型的同步定位与地图构建方法及装置，将WIFI指纹技术与基于啮齿类动物模型的仿生定位技术相融合，实现对位姿细胞网络的修正，以获得最优路径经历图，继而实现对机器人的进行精确定位。The technical problem to be solved by the present invention is to provide a method and device for synchronous positioning and map construction based on a rodent model, which integrates WIFI fingerprint technology with bionic positioning technology based on a rodent model to realize the control of the pose cell network. Correction to obtain the optimal path experience map, and then realize the precise positioning of the robot.
为解决上述技术问题，本发明提供的技术方案是：In order to solve the problems of the technologies described above, the technical solution provided by the invention is:
第一方面，本发明提供一种基于啮齿类动物模型的同步定位与地图构建方法，方法包括：In a first aspect, the present invention provides a method for synchronous positioning and map construction based on a rodent model, the method comprising:
获取机器人的当前视觉场景图像信息；Obtain the current visual scene image information of the robot;
根据预先构建的啮齿类动物模型，从预先构建的视觉信息库中匹配出与当前视觉场景图像信息具有最大场景相似度的位姿信息；According to the pre-built rodent model, match the pose information with the maximum scene similarity with the current visual scene image information from the pre-built visual information library;
在场景相似度低于设定阈值时，获取机器人的当前WIFI信号强度集合；When the scene similarity is lower than the set threshold, obtain the current WIFI signal strength set of the robot;
根据啮齿类动物模型，从预先构建的WIFI指纹地图中匹配出与当前WIFI信号强度集合具有最大指纹相似度的位姿信息；According to the rodent model, match the pose information with the largest fingerprint similarity with the current WIFI signal strength set from the pre-built WIFI fingerprint map;
根据最大场景相似度和/最大指纹相似度对应的位姿信息对机器人进行同步定位与地图构建。According to the pose information corresponding to the maximum scene similarity and/or maximum fingerprint similarity, the robot is synchronously positioned and map-built.
进一步地，方法还包括：Further, the method also includes:
根据最大场景相似度对视觉信息库进行更新；Update the visual information base according to the maximum scene similarity;
根据最大指纹相似度对WIFI指纹地图进行更新。The WIFI fingerprint map is updated according to the maximum fingerprint similarity.
进一步地，获取机器人的当前WIFI信号强度集合，包括：Further, obtain the current WIFI signal strength set of the robot, including:
确定有效的无线接入点；Identify valid wireless access points;
接收每个有效的无线接入点在当前位置产生的当前接收信号强度均值；Receive the average value of the current received signal strength generated by each valid wireless access point at the current location;
将所有有效的无线接入点对应的当前接收信号强度均值确定为机器人的当前WIFI信号强度集合。The mean value of the current received signal strength corresponding to all valid wireless access points is determined as the current WIFI signal strength set of the robot.
进一步地，确定有效的无线接入点，包括：Further, determining a valid wireless access point includes:
确定有效的无线接入点的数量；Determining the number of active wireless access points;
从所有无线接入点组成的无线接入点集合中随机选定两个无线接入点作为参考接入点；Randomly select two wireless access points from a wireless access point set composed of all wireless access points as reference access points;
计算两个参考接入点间的第一互信息；calculating first mutual information between two reference access points;
从无线接入点集合中获取可使第二互信息最小的无线接入点；Acquire a wireless access point that can minimize the second mutual information from the wireless access point set;
从无线接入点集合中获取可使第三互信息最小的无线接入点；Acquire a wireless access point that can minimize the third mutual information from the wireless access point set;
以此类推，直至获取到足够数量的无线接入点。And so on, until a sufficient number of wireless access points are obtained.
进一步地，根据啮齿类动物模型，从预先构建的WIFI指纹地图中匹配出与当前WIFI信号强度集合具有最大指纹相似度的位姿信息，包括：Further, according to the rodent model, the pose information with the largest fingerprint similarity to the current WIFI signal strength set is matched from the pre-built WIFI fingerprint map, including:
根据当前WIFI信号强度集合，以及根据预先构建的贝叶斯后验估计模型获取机器人的估计位置；Obtain the estimated position of the robot according to the current WIFI signal strength set and the pre-built Bayesian posterior estimation model;
根据估计位置，WIFI指纹地图，以及根据啮齿类动物模型中的位姿细胞网络，进行位姿信息匹配，以匹配出与当前WIFI信号强度集合具有最大指纹相似度的位姿信息。According to the estimated position, the WIFI fingerprint map, and the pose cell network in the rodent model, the pose information is matched to match the pose information with the maximum fingerprint similarity to the current WIFI signal strength set.
进一步地，位姿细胞网络进行位姿信息匹配，包括：Further, the pose cell network performs pose information matching, including:
根据估计位置，从WIFI指纹地图中提取出与估计位置相邻的至少一个经历单元；Extracting at least one experience unit adjacent to the estimated position from the WIFI fingerprint map according to the estimated position;
计算每个经历单元的WIFI信号强度与当前WIFI信号强度间的欧氏距离；Calculate the Euclidean distance between the WIFI signal strength of each experience unit and the current WIFI signal strength;
获取最大欧式距离对应的经历单元所指向的位姿信息，并将获取到的位姿信息确定为机器人的当前位姿信息。Obtain the pose information pointed to by the experienced unit corresponding to the maximum Euclidean distance, and determine the obtained pose information as the current pose information of the robot.
进一步地，欧式距离的计算公式为：Further, the formula for calculating the Euclidean distance is:
其中，(xpc,ypc,θpc)为经历单元对应的位姿细胞坐标；(xi,yi,θi)为与当前位置对应的位姿细胞坐标；ra为(x,y)平面的区域常数，θa为θ维上的区域常数。Among them, (xpc ,ypc ,θpc ) are the coordinates of the pose cell corresponding to the experienced unit; (xi ,yi ,θi ) are the coordinates of the pose cell corresponding to the current position; ra is (x,y ) plane area constant, θa is the area constant on the θ dimension.
进一步地，WIFI指纹地图的构建，包括：Further, the construction of the WIFI fingerprint map includes:
选定参考点；selected reference point;
在每个参考点上，对每个预设的无线接入点的信号强度进行连续采样后求平均，以获取每个无线接入点在参考点处的接收信号强度均值；At each reference point, the signal strength of each preset wireless access point is averaged after continuous sampling, so as to obtain the mean value of the received signal strength of each wireless access point at the reference point;
根据各参考点对应的所有无线接入点的接收信号强度均值，按照预设规则进行构建WIFI指纹地图并存储。According to the average received signal strength of all wireless access points corresponding to each reference point, a WIFI fingerprint map is constructed and stored according to preset rules.
进一步地，WIFI指纹地图的数据存储结构为：Further, the data storage structure of the WIFI fingerprint map is:
IM＝{φ,A,M,MACi}；其中，IM={φ,A,M,MACi }; where,
φ＝{L1,L2,…,Li,…,Lk}；A＝{AP1,AP2,…,APi,…,APR}；φ={L1 ,L2 ,...,Li ,...,Lk }; A={AP1 ,AP2 ,...,APi ,...,APR };
MACi表示第i个参考点的MAC地址值；MACi represents the MAC address value of the i-th reference point;
其中，IM表示WIFI指纹地图；Li＝(xi,yi)表示第i个参考点的位置，k为参考点的数量，φ表示所有参考点的位置集合，表示地图中所有观测到的无线接入点的组成集合，R为观测到的无线接入点的数量，M为各参考点对应各无线接入点的接收信号强度均值的集合，其中为第R个无线接入点在参考点Lk处的接收信号强度均值。Among them, IM represents the WIFI fingerprint map; Li = (xi , yi ) represents the position of the i-th reference point, k is the number of reference points, φ represents the set of positions of all reference points, and represents all observed points in the map A set of wireless access points, R is the number of observed wireless access points, M is the set of average received signal strengths of each reference point corresponding to each wireless access point, where is the average received signal strength of the Rth wireless access point at the reference point Lk .
第二方面，本发明提供一种基于啮齿类动物模型的机器人同步定位与地图构建装置，装置包括：In a second aspect, the present invention provides a device for synchronous positioning and map construction of a robot based on a rodent model, the device comprising:
信息获取单元，用于获取机器人的当前视觉场景图像信息；An information acquisition unit, configured to acquire the current visual scene image information of the robot;
第一匹配单元，用于根据预先构建的啮齿类动物模型，从预先构建的视觉信息库中匹配出与当前视觉场景图像信息具有最大场景相似度的位姿信息；The first matching unit is used to match the pose information with the maximum scene similarity with the current visual scene image information from the pre-built visual information library according to the pre-built rodent model;
数据判断单元，用于在场景相似度低于设定阈值时，获取机器人的当前WIFI信号强度集合；The data judging unit is used to obtain the current WIFI signal strength set of the robot when the scene similarity is lower than the set threshold;
第二匹配单元，用于根据啮齿类动物模型，从预先构建的WIFI指纹地图中匹配出与当前WIFI信号强度集合具有最大指纹相似度的位姿信息；The second matching unit is used to match the pose information with the maximum fingerprint similarity with the current WIFI signal strength set from the pre-built WIFI fingerprint map according to the rodent model;
同步定位单元，用于根据最大场景相似度和/最大指纹相似度对应的位姿信息对机器人进行同步定位与地图构建。The synchronous positioning unit is used for synchronous positioning and map construction of the robot according to the pose information corresponding to the maximum scene similarity and/or maximum fingerprint similarity.
本发明提供的基于啮齿类动物模型的同步定位与地图构建方法及装置，将WIFI指纹技术与基于啮齿类动物模型的仿生定位技术相融合，实现对位姿细胞网络的修正，以获得最优路径经历图，继而实现对机器人的进行精确定位。The rodent model-based synchronous positioning and map construction method and device provided by the present invention combine the WIFI fingerprint technology with the rodent model-based bionic positioning technology to realize the correction of the pose cell network to obtain the optimal path Through the map, and then realize the precise positioning of the robot.
附图说明Description of drawings
图1是本发明实施例提供的方法的流程图；Fig. 1 is the flowchart of the method provided by the embodiment of the present invention;
图2是本发明实施例提供的装置的框图；Fig. 2 is a block diagram of a device provided by an embodiment of the present invention;
图3是本发明实施例提供的利用WIFI指纹匹配过程关系示意图。Fig. 3 is a schematic diagram of the process of fingerprint matching using WIFI provided by the embodiment of the present invention.
具体实施方式Detailed ways
下面通过具体的实施例进一步说明本发明，但是，应当理解为，这些实施例仅仅是用于更详细具体地说明之用，而不应理解为用于以任何形式限制本发明。The present invention will be further illustrated by specific examples below, but it should be understood that these examples are only used for more detailed description, and should not be construed as limiting the present invention in any form.
实施例一Embodiment one
结合图1，本实施例提供的基于啮齿类动物模型的同步定位与地图构建方法，方法包括：In conjunction with Fig. 1, the method for synchronous positioning and map construction based on the rodent model provided in this embodiment includes:
步骤S1，获取机器人的当前视觉场景图像信息；Step S1, acquiring the current visual scene image information of the robot;
步骤S2，根据预先构建的啮齿类动物模型，从预先构建的视觉信息库中匹配出与当前视觉场景图像信息具有最大场景相似度的位姿信息；Step S2, according to the pre-built rodent model, match the pose information with the maximum scene similarity with the current visual scene image information from the pre-built visual information database;
步骤S3，在场景相似度低于设定阈值时，获取机器人的当前WIFI信号强度集合；Step S3, when the scene similarity is lower than the set threshold, obtain the current WIFI signal strength set of the robot;
步骤S4，根据啮齿类动物模型，从预先构建的WIFI指纹地图中匹配出与当前WIFI信号强度集合具有最大指纹相似度的位姿信息；Step S4, according to the rodent model, match the pose information with the maximum fingerprint similarity with the current WIFI signal strength set from the pre-built WIFI fingerprint map;
步骤S5，根据最大场景相似度和/最大指纹相似度对应的位姿信息对机器人进行同步定位与地图构建。Step S5, perform synchronous positioning and map construction on the robot according to the pose information corresponding to the maximum scene similarity and/or maximum fingerprint similarity.
本发明实施例提供的基于啮齿类动物模型的同步定位与地图构建方法，将WIFI指纹技术与基于啮齿类动物模型的仿生定位技术相融合，实现对位姿细胞网络的修正，以获得最优路径经历图，继而实现对机器人的进行精确定位。The rodent model-based synchronous positioning and map construction method provided by the embodiment of the present invention combines the WIFI fingerprint technology with the rodent model-based bionic positioning technology to realize the correction of the pose cell network to obtain the optimal path Through the map, and then realize the precise positioning of the robot.
优选地，方法还包括：Preferably, the method also includes:
根据最大场景相似度对视觉信息库进行更新；Update the visual information base according to the maximum scene similarity;
根据最大指纹相似度对WIFI指纹地图进行更新。The WIFI fingerprint map is updated according to the maximum fingerprint similarity.
本实施例中，具体地，最大场景相似度为R1，最大指纹相似度为R2，λ1，λ2为预先设定的两个场景相似度阈值，ε1，ε2为预先设定的两个指纹相似度阈值，ε1＜ε2。更加具体地，当R1＞λ2时，匹配成功，此时不需要进行WIFI指纹匹配，不对视觉信息库进行更新；当R1＜λ1时，匹配失败，对视觉信息库进行更新；λ1＜R1＜λ2且R2＞ε2时，匹配成功，不进行更新；当λ1＜R1＜λ2且ε1＜R2＜ε2时，匹配成功，不进行更新；当λ1＜R1＜λ2且R2＜ε1时，匹配失败，对WIFI指纹地图进行更新。In this embodiment, specifically, the maximum scene similarity is R1, the maximum fingerprint similarity is R2, λ1 and λ2 are two preset scene similarity thresholds, ε1 and ε2 are two preset thresholds fingerprint similarity threshold, ε1 <ε2 . More specifically, when R1 >λ2 , the matching is successful, WIFI fingerprint matching is not needed at this time, and the visual information database is not updated; when R1 <λ1 , the matching fails, and the visual information database is updated; λ1 <R1 <λ2 and R2 >ε2 , the matching is successful and no update is performed; when λ1 <R1 <λ2 and ε1 <R2 <ε2 , the matching is successful and no update is performed; When λ1 <R1 <λ2 and R2 <ε1 , the matching fails, and the WIFI fingerprint map is updated.
进一步优选地，获取机器人的当前WIFI信号强度集合，包括：Further preferably, the current WIFI signal strength set of the robot is obtained, including:
确定有效的无线接入点；Identify valid wireless access points;
接收每个有效的无线接入点在当前位置产生的当前接收信号强度均值；Receive the average value of the current received signal strength generated by each valid wireless access point at the current location;
将所有有效的无线接入点对应的当前接收信号强度均值确定为机器人的当前WIFI信号强度集合。The mean value of the current received signal strength corresponding to all valid wireless access points is determined as the current WIFI signal strength set of the robot.
本实施例中，可以结合实际需要对有效的无线接入点(Access Point,AP)的数量进行设定，如此，可使信号空间的维度降低，减少计算量。具体地，本实施例中，例如，某室内环境下，具有T个可用的AP，需要从中选取最佳的S个有效的AP，在确定了有效的AP后，依次测量各参考点来自不同AP的接收信号强度值(Received Signal Strength Indication，RSSI)作为该参考点AP的信号特征数，并按一定格式记录在位置指纹数据库中，该数据库也被称为位置指纹地图或WIFI指纹图。In this embodiment, the number of effective wireless access points (Access Points, APs) can be set according to actual needs. In this way, the dimension of the signal space can be reduced and the amount of calculation can be reduced. Specifically, in this embodiment, for example, in an indoor environment, there are T available APs, and it is necessary to select the best S valid APs among them. The received signal strength value (Received Signal Strength Indication, RSSI) of the reference point AP is used as the signal characteristic number, and is recorded in the location fingerprint database according to a certain format. This database is also called a location fingerprint map or a WIFI fingerprint map.
具体地，WIFI指纹地图的构建，包括：Specifically, the construction of the WIFI fingerprint map includes:
选定参考点；selected reference point;
在每个参考点上，对每个预设的无线接入点的信号强度进行连续采样后求平均，以获取每个无线接入点在参考点处的接收信号强度均值；At each reference point, the signal strength of each preset wireless access point is averaged after continuous sampling, so as to obtain the mean value of the received signal strength of each wireless access point at the reference point;
根据各参考点对应的所有无线接入点的接收信号强度均值，按照预设规则进行构建WIFI指纹地图并存储。According to the average received signal strength of all wireless access points corresponding to each reference point, a WIFI fingerprint map is constructed and stored according to preset rules.
进一步具体地，WIFI指纹地图的数据存储结构为：Further specifically, the data storage structure of the WIFI fingerprint map is:
IM＝{φ,A,M,MACi}；其中，IM={φ,A,M,MACi }; where,
φ＝{L1,L2,…,Li,…,Lk}；A＝{AP1,AP2,…,APi,…,APR}；φ={L1 ,L2 ,...,Li ,...,Lk }; A={AP1 ,AP2 ,...,APi ,...,APR };
MACi表示第i个参考点的MAC地址值；MACi represents the MAC address value of the i-th reference point;
其中，IM表示WIFI指纹地图；Li＝(xi,yi)表示第i个参考点的位置，k为参考点的数量，φ表示所有参考点的位置集合，表示地图中所有观测到的无线接入点的组成集合，R为观测到的无线接入点的数量，M为各参考点对应各无线接入点的接收信号强度均值的集合，其中为第R个无线接入点在参考点Lk处的接收信号强度均值。Among them, IM represents the WIFI fingerprint map; Li = (xi , yi ) represents the position of the i-th reference point, k is the number of reference points, φ represents the set of positions of all reference points, and represents all observed points in the map A set of wireless access points, R is the number of observed wireless access points, M is the set of average received signal strengths of each reference point corresponding to each wireless access point, where is the average received signal strength of the Rth wireless access point at the reference point Lk .
本实施例中，室内环境下WIFI指纹图的建立方法如下，在定位环境中按照一定规则选取参考点，并在每个参考点上对AP的信号强度连续采样一段时间，得到每个AP的均值存入数据库中，构成定位指纹图IM，其中，其中，Li＝(xi,yi)表示参考点的位置,k为参考点的数量，表示所有参考点的位置集合；A＝{AP1,AP2,...,APR}表示地图中所有观测到AP的组成集合。In this embodiment, the establishment method of the WIFI fingerprint in the indoor environment is as follows. In the positioning environment, reference points are selected according to certain rules, and the signal strength of the AP is continuously sampled at each reference point for a period of time to obtain the average value of each AP. Stored in the database to form a positioning fingerprint IM, wherein, in, Li ＝(xi , yi ) indicates the position of the reference point, k is the number of reference points, Indicates the location set of all reference points; A={AP1 , AP2 ,...,APR } indicates the composition set of all observed APs in the map.
M为定位指纹中所有均值的集合，其中为第j个AP在参考点Li处的均值,MACi表示第i个参考点的MAC地址值，具体地，M的数据结构如公式(1)所示。M is the set of all means in the positioning fingerprint, where is the mean value of the j-th AP at the reference point Li , and MACi represents the MAC address value of the i-th reference point, specifically, the data structure of M is shown in formula (1).
更加具体地，确定有效的无线接入点，包括：More specifically, identifying valid wireless access points includes:
确定有效的无线接入点的数量；Determining the number of active wireless access points;
从所有无线接入点组成的无线接入点集合中随机选定两个无线接入点作为参考接入点；Randomly select two wireless access points from a wireless access point set composed of all wireless access points as reference access points;
计算两个参考接入点间的第一互信息；calculating first mutual information between two reference access points;
从无线接入点集合中获取可使第二互信息最小的无线接入点；Acquire a wireless access point that can minimize the second mutual information from the wireless access point set;
从无线接入点集合中获取可使第三互信息最小的无线接入点；Acquire a wireless access point that can minimize the third mutual information from the wireless access point set;
以此类推，直至获取到足够数量的无线接入点。And so on, until a sufficient number of wireless access points are obtained.
本实施例中，更加具体地，设定室内定位环境可用的参考点AP个数为T，选取其中S个AP的优化子集则可以将信号空间的维度从T维降到S维，因而可以减少计算量。具体地，本实施例采用互信息最小化AP选取策略，且具体步骤如下：In this embodiment, more specifically, the number of reference point APs available in the indoor positioning environment is set as T, and selecting an optimal subset of S APs among them can reduce the dimension of the signal space from the T dimension to the S dimension, so that Reduce the amount of computation. Specifically, this embodiment adopts the mutual information minimization AP selection strategy, and the specific steps are as follows:
1)对于选取的S个AP进行两两组合，按照下式计算每个组合的互信息，查找出互信息最小的组合，其对应的APm，APn作为两个初始参考点AP；1) Combining the selected S APs in pairs, calculate the mutual information of each combination according to the following formula, find out the combination with the minimum mutual information, and use its corresponding APm and APn as the two initial reference points AP;
MI(APm,APn)＝H(APm)+H(APn)-H(APm,APn) (2)MI(APm ,APn )=H(APm )+H(APn )-H(APm ,APn ) (2)
公式(2)中：MI(APm,APn)表示两个不同AP的互信息，即，第一互信息，H(APm,APn)表示两个AP的组合信息熵。In the formula (2): MI(APm , APn ) represents the mutual information of two different APs, that is, the first mutual information, and H(APm , APn ) represents the combined information entropy of the two APs.
2)按照公式(3)计算某AP与两个初始AP组合的互信息。2) Calculate the mutual information between an AP and two initial AP combinations according to formula (3).
MI(APm,APn,APi)＝H(APm,APn)+H(APi)-H(APm,APn,APi) (3)MI(APm ,APn ,APi )=H(APm ,APn )+H(APi )-H(APm ,APn ,APi ) (3)
找出能使得MI最小的AP作为最优化AP子集的第三个AP。Find the AP that can make MI the smallest as the third AP of the optimal AP subset.
3)依次按照第2)步的形式选取下一个最优的AP，依次迭代，直到选取出S个最优AP为止。需要说明的是，第R个最优的AP的选取算式如下：3) Select the next optimal AP sequentially according to the form of step 2), and iterate successively until S optimal APs are selected. It should be noted that the formula for selecting the Rth optimal AP is as follows:
MI(AP1,AP2,…,APR)＝H(AP1,AP2,…,APR-1)+H(APR)-H(APm,APn,…,APR) (4)MI(AP1 ,AP2 ,…,APR )=H(AP1 ,AP2 ,…,APR-1 )+H(APR )-H(APm ,APn ,…,APR ) ( 4)
优选地，根据啮齿类动物模型，从预先构建的WIFI指纹地图中匹配出与当前WIFI信号强度集合具有最大指纹相似度的位姿信息，包括：Preferably, according to the rodent model, the pose information having the maximum fingerprint similarity with the current WIFI signal strength set is matched from the pre-built WIFI fingerprint map, including:
根据当前WIFI信号强度集合，以及根据预先构建的贝叶斯后验估计模型获取机器人的估计位置；Obtain the estimated position of the robot according to the current WIFI signal strength set and the pre-built Bayesian posterior estimation model;
根据估计位置，WIFI指纹地图，以及根据啮齿类动物模型中的位姿细胞网络，进行位姿信息匹配，以匹配出与当前WIFI信号强度集合具有最大指纹相似度的位姿信息。According to the estimated position, the WIFI fingerprint map, and the pose cell network in the rodent model, the pose information is matched to match the pose information with the maximum fingerprint similarity to the current WIFI signal strength set.
本实施例中，采用贝叶斯位置估计策略，且具体地，针对上述的互信息最小化AP选取策略，进一步使用贝叶斯后验估计进行组合优化，使得WIFI指纹定位算法的位置估计精度和可靠度大大提升。In this embodiment, the Bayesian position estimation strategy is adopted, and specifically, for the above-mentioned mutual information minimization AP selection strategy, Bayesian posterior estimation is further used for combined optimization, so that the position estimation accuracy of the WIFI fingerprint positioning algorithm and The reliability is greatly improved.
贝叶斯后验估计的基本原理如下式所示：The basic principle of Bayesian posterior estimation is as follows:
式中：RSSI表示多个AP在位置估计点的RSSI观测值；p(Li|RSSI)表示位置Li的在给定RSSI下的条件概率，即在观测到RSSI向量的情况下，定位点出现在Li的概率；p(RSSI|Li)表示位置Li的概率；p(Li)表示位置Li的概率，通常不考虑指纹点之间的差异，即指纹点等概率；p(RSSI)表示RSSI出现的全概率，其算式如下：In the formula: RSSI indicates the RSSI observation value of multiple APs at the location estimation point; p(Li |RSSI) indicates the conditional probability of the location Li under the given RSSI, that is, when the RSSI vector is observed, the location point The probability of appearing in Li ; p(RSSI|Li ) indicates the probability of position Li ; p(Li ) indicates the probability of position Li , usually regardless of the difference between fingerprint points, that is, the equal probability of fingerprint points; p (RSSI) indicates the total probability of RSSI occurrence, and its formula is as follows:
其中，C(RSSI1,RSSI2,…,RSSIM)表示指纹点观测到的指定RSSI向量的个数；K表示指纹点观测历元数。Among them, C(RSSI1, RSSI2,...,RSSIM) represents the number of specified RSSI vectors observed by fingerprint points; K represents the number of epochs observed by fingerprint points.
将上述全概率算式回代至贝叶斯后验估计式，从而计算出后验条件概率。采用多个指纹点的贝叶斯权重位置估计算式能够在较短时间内算出位置估计点的位置，令估计点的位置为p，则估计位置的计算公式如下：Substitute the above full probability formula back into the Bayesian posterior estimation formula to calculate the posterior conditional probability. The Bayesian weight position estimation formula using multiple fingerprint points can calculate the position of the position estimation point in a relatively short period of time. Let the position of the estimated point be p, then the calculation formula of the estimated position is as follows:
式中：(x,y)表示位置估计点的二维坐标，(xi,yi)表示第i个指纹点的坐标，ωi表征第i个指纹点的加权权重，即为贝叶斯后验条件的概率，K表示邻近点个数In the formula: (x, y) represents the two-dimensional coordinates of the position estimation point, (xi , yi ) represents the coordinates of the i-th fingerprint point, ωi represents the weighted weight of the i-th fingerprint point, which is the Bayesian The probability of the posterior condition, K represents the number of adjacent points
优选地，位姿细胞网络进行位姿信息匹配，包括：Preferably, the pose cell network performs pose information matching, including:
根据估计位置，从WIFI指纹地图中提取出与估计位置相邻的至少一个经历单元；Extracting at least one experience unit adjacent to the estimated position from the WIFI fingerprint map according to the estimated position;
计算每个经历单元的WIFI信号强度与当前WIFI信号强度间的欧氏距离；Calculate the Euclidean distance between the WIFI signal strength of each experience unit and the current WIFI signal strength;
获取最大欧式距离对应的经历单元所指向的位姿信息，并将获取到的位姿信息确定为机器人的当前位姿信息。Obtain the pose information pointed to by the experienced unit corresponding to the maximum Euclidean distance, and determine the obtained pose information as the current pose information of the robot.
进一步优选地，欧式距离的计算公式为：Further preferably, the formula for calculating the Euclidean distance is:
其中，(xpc,ypc,θpc)为经历单元对应的位姿细胞坐标；(xi,yi,θi)为与当前位置对应的位姿细胞坐标；ra为(x,y)平面的区域常数，θa为θ维上的区域常数。Among them, (xpc ,ypc ,θpc ) are the coordinates of the pose cell corresponding to the experienced unit; (xi ,yi ,θi ) are the coordinates of the pose cell corresponding to the current position; ra is (x,y ) plane area constant, θa is the area constant on the θ dimension.
本实施例中，每个经历单元都具有一个活性水平，活性水平由位姿感知细胞和WIFI指纹中能量峰与每一个经历单元之间接近程度所决定。每一个经历在位姿感知细胞和WIFI指纹中有一个相关的活性区域。当能量峰处于这些活性区域时，该激励立刻被激活，这些区域在位姿感知细胞内部是连续的，而在WIFI指纹中的相关区域却是非连续的。每个经历ei由经历活性水平Ei，WIFI信号强度Ri所决定。其中，ei＝{Ei,Ri}，In this embodiment, each experience unit has an activity level, and the activity level is determined by the proximity between the pose sensing cells and the energy peak in the WIFI fingerprint and each experience unit. Each experience has an associated active area in the pose sensing cells and WIFI fingerprints. When the energy peak is in these active areas, the excitation is activated immediately, and these areas are continuous inside the pose sensing cells, but the relevant areas in the WIFI fingerprint are discontinuous. Each experience ei is determined by the experience activity level Ei and the WIFI signal strength Ri . Among them, ei = {Ei , Ri },
一个经历单元的能级Exyθ和第i个经历单元的总能级水平Ei由公式(9)和公式(10)计算可得。The energy level Exyθ of an experienced unit and the total energy level Ei of the i-th experienced unit can be calculated by formula (9) and formula (10).
其中，xpcypc和θpc为最大活性姿态细胞的坐标；xi、yi、θi为与该经历相关的位姿感知细胞的坐标；ra为(x,y)平面的区域常数；θa为θ维上的区域常数。Rcurr为当前WIFI信号强度；Ri为与经历i相关的WIFI信号强度。Among them, xpc ypc and θpc are the coordinates of the most active attitude cell; xi , yi , θi are the coordinates of the pose sensing cells related to the experience; ra is the area constant of the (x,y) plane ; θa is the area constant on the θ dimension. Rcurr is the current WIFI signal strength; Ri is the WIFI signal strength related to experience i.
需要说明的是，本实施例中，如图3所示地，利用WIFI指纹匹配过程关系示意图。将无线信号网络WIFI作为一种传感器用到啮齿类动物模型当中，其定位模型有三个主要部分组成，分别为WIFI指纹，位姿细胞网络和经历图。WIFI指纹获取环境的WIFI信号强度，被称为WIFI信号强度模板。WIFI指纹信息用来辨识熟悉的环境。当新输入的WIFI信号强度信息与已存在的WIFI信号强度模板匹配时，位姿细胞网络的活性因子被激活，二者结合能够很大程度上阻止错误匹配的发生，产生更为准确的经历图。It should be noted that, in this embodiment, as shown in FIG. 3 , it is a schematic diagram of a process relationship using WIFI fingerprint matching. The wireless signal network WIFI is used as a sensor in the rodent model. The positioning model consists of three main parts, namely the WIFI fingerprint, the pose cell network and the experience map. The WIFI fingerprint acquires the WIFI signal strength of the environment, which is called a WIFI signal strength template. WIFI fingerprint information is used to identify familiar environments. When the newly input WIFI signal strength information matches the existing WIFI signal strength template, the active factors of the pose cell network are activated. The combination of the two can largely prevent the occurrence of wrong matching and generate a more accurate experience map. .
实施例二Embodiment two
结合图2，本发明实施例提供的基于啮齿类动物模型的机器人同步定位与地图构建装置，装置包括：In conjunction with FIG. 2, the rodent model-based robot synchronous positioning and map construction device provided by the embodiment of the present invention includes:
信息获取单元1，用于获取机器人的当前视觉场景图像信息；An information acquisition unit 1, configured to acquire the current visual scene image information of the robot;
第一匹配单元2，用于根据预先构建的啮齿类动物模型，从预先构建的视觉信息库中匹配出与当前视觉场景图像信息具有最大场景相似度的位姿信息；The first matching unit 2 is used to match the pose information with the maximum scene similarity with the current visual scene image information from the pre-built visual information library according to the pre-built rodent model;
数据判断单元3，用于在场景相似度低于设定阈值时，获取机器人的当前WIFI信号强度集合；The data judging unit 3 is used to obtain the current WIFI signal strength set of the robot when the scene similarity is lower than the set threshold;
第二匹配单元4，用于根据啮齿类动物模型，从预先构建的WIFI指纹地图中匹配出与当前WIFI信号强度集合具有最大指纹相似度的位姿信息；The second matching unit 4 is used to match the pose information with the maximum fingerprint similarity with the current WIFI signal strength set from the pre-built WIFI fingerprint map according to the rodent model;
同步定位单元5，用于根据最大场景相似度和/最大指纹相似度对应的位姿信息对机器人进行同步定位与地图构建。The synchronous positioning unit 5 is used to perform synchronous positioning and map construction on the robot according to the pose information corresponding to the maximum scene similarity and/or maximum fingerprint similarity.
本发明实施例提供的基于啮齿类动物模型的同步定位与地图构建装置，将WIFI指纹技术与基于啮齿类动物模型的仿生定位技术相融合，实现对位姿细胞网络的修正，以获得最优路径经历图，继而实现对机器人的进行精确定位。The rodent model-based synchronous positioning and map construction device provided by the embodiment of the present invention combines WIFI fingerprint technology with the rodent model-based bionic positioning technology to realize the correction of the pose cell network to obtain the optimal path Through the map, and then realize the precise positioning of the robot.
优选地，装置还包括数据更新单元，用于Preferably, the device further includes a data updating unit for
根据最大场景相似度对视觉信息库进行更新；Update the visual information base according to the maximum scene similarity;
根据最大指纹相似度对WIFI指纹地图进行更新。The WIFI fingerprint map is updated according to the maximum fingerprint similarity.
本实施例中，具体地，最大场景相似度为R1，最大指纹相似度为R2，λ1，λ2为预先设定的两个场景相似度阈值，ε1，ε2为预先设定的两个指纹相似度阈值，ε1＜ε2。更加具体地，当R1＞λ2时，匹配成功，此时不需要进行WIFI指纹匹配，不对视觉信息库进行更新；当R1＜λ1时，匹配失败，对视觉信息库进行更新；λ1＜R1＜λ2且R2＞ε2时，匹配成功，不进行更新；当λ1＜R1＜λ2且ε1＜R2＜ε2时，匹配成功，不进行更新；当λ1＜R1＜λ2且R2＜ε1时，匹配失败，对WIFI指纹地图进行更新。In this embodiment, specifically, the maximum scene similarity is R1, the maximum fingerprint similarity is R2, λ1 and λ2 are two preset scene similarity thresholds, ε1 and ε2 are two preset thresholds fingerprint similarity threshold, ε1 <ε2 . More specifically, when R1 >λ2 , the matching is successful, WIFI fingerprint matching is not needed at this time, and the visual information database is not updated; when R1 <λ1 , the matching fails, and the visual information database is updated; λ1 <R1 <λ2 and R2 >ε2 , the matching is successful and no update is performed; when λ1 <R1 <λ2 and ε1 <R2 <ε2 , the matching is successful and no update is performed; When λ1 <R1 <λ2 and R2 <ε1 , the matching fails, and the WIFI fingerprint map is updated.
进一步优选地，获取机器人的当前WIFI信号强度集合，包括：Further preferably, the current WIFI signal strength set of the robot is obtained, including:
确定有效的无线接入点；Identify valid wireless access points;
接收每个有效的无线接入点在当前位置产生的当前接收信号强度均值；Receive the average value of the current received signal strength generated by each valid wireless access point at the current location;
将所有有效的无线接入点对应的当前接收信号强度均值确定为机器人的当前WIFI信号强度集合。The mean value of the current received signal strength corresponding to all valid wireless access points is determined as the current WIFI signal strength set of the robot.
本实施例中，可以结合实际需要对有效的无线接入点(Access Point,AP)的数量进行设定，如此，可使信号空间的维度降低，减少计算量。具体地，本实施例中，例如，某室内环境下，具有T个可用的AP，需要从中选取最佳的S个有效的AP，在确定了有效的AP后，依次测量各参考点来自不同AP的接收信号强度值(Received Signal Strength Indication，RSSI)作为该参考点AP的信号特征数，并按一定格式记录在位置指纹数据库中，该数据库也被称为位置指纹地图或WIFI指纹图。In this embodiment, the number of effective wireless access points (Access Points, APs) can be set according to actual needs. In this way, the dimension of the signal space can be reduced and the amount of calculation can be reduced. Specifically, in this embodiment, for example, in an indoor environment, there are T available APs, and it is necessary to select the best S effective APs among them. The received signal strength value (Received Signal Strength Indication, RSSI) of the reference point AP is used as the signal characteristic number, and is recorded in a location fingerprint database in a certain format, which is also called a location fingerprint map or a WIFI fingerprint map.
具体地，WIFI指纹地图的构建，包括：Specifically, the construction of the WIFI fingerprint map includes:
选定参考点；selected reference point;
在每个参考点上，对每个预设的无线接入点的信号强度进行连续采样后求平均，以获取每个无线接入点在参考点处的接收信号强度均值；At each reference point, the signal strength of each preset wireless access point is averaged after continuous sampling, so as to obtain the mean value of the received signal strength of each wireless access point at the reference point;
根据各参考点对应的所有无线接入点的接收信号强度均值，按照预设规则进行构建WIFI指纹地图并存储。According to the average received signal strength of all wireless access points corresponding to each reference point, a WIFI fingerprint map is constructed and stored according to preset rules.
进一步具体地，WIFI指纹地图的数据存储结构为：Further specifically, the data storage structure of the WIFI fingerprint map is:
IM＝{φ,A,M,MACi}；其中，IM={φ,A,M,MACi }; where,
φ＝{L1,L2,…,Li,…,Lk}；A＝{AP1,AP2,…,APi,…,APR}；φ={L1 ,L2 ,...,Li ,...,Lk }; A={AP1 ,AP2 ,...,APi ,...,APR };
MACi表示第i个参考点的MAC地址值；MACi represents the MAC address value of the i-th reference point;
其中，IM表示WIFI指纹地图；Li＝(xi,yi)表示第i个参考点的位置，k为参考点的数量，φ表示所有参考点的位置集合，表示地图中所有观测到的无线接入点的组成集合，R为观测到的无线接入点的数量，M为各参考点对应各无线接入点的接收信号强度均值的集合，其中为第R个无线接入点在参考点Lk处的接收信号强度均值。Among them, IM represents the WIFI fingerprint map; Li = (xi , yi ) represents the position of the i-th reference point, k is the number of reference points, φ represents the set of positions of all reference points, and represents all observed points in the map A set of wireless access points, R is the number of observed wireless access points, M is the set of average received signal strengths of each reference point corresponding to each wireless access point, where is the average received signal strength of the Rth wireless access point at the reference point Lk .
本实施例中，室内环境下WIFI指纹图的建立方法如下，在定位环境中按照一定规则选取参考点，并在每个参考点上对AP的信号强度连续采样一段时间，得到每个AP的均值存入数据库中，构成定位指纹图IM，其中，其中，Li＝(xi,yi)表示参考点的位置,k为参考点的数量，表示所有参考点的位置集合；A＝{AP1,AP2,...,APR}表示地图中所有观测到AP的组成集合。In this embodiment, the establishment method of the WIFI fingerprint in the indoor environment is as follows. In the positioning environment, reference points are selected according to certain rules, and the signal strength of the AP is continuously sampled at each reference point for a period of time to obtain the average value of each AP. Stored in the database to form a positioning fingerprint IM, wherein, in, Li ＝(xi , yi ) indicates the position of the reference point, k is the number of reference points, Indicates the location set of all reference points; A={AP1 , AP2 ,...,APR } indicates the composition set of all observed APs in the map.
M为定位指纹中所有均值的集合，其中为第j个AP在参考点Li处的均值,MACi表示第i个参考点的MAC地址值，具体地，M的数据结构如公式(1)所示。M is the set of all means in the positioning fingerprint, where is the mean value of the j-th AP at the reference point Li , and MACi represents the MAC address value of the i-th reference point, specifically, the data structure of M is shown in formula (1).
更加具体地，确定有效的无线接入点，包括：More specifically, identifying valid wireless access points includes:
确定有效的无线接入点的数量；Determining the number of active wireless access points;
从所有无线接入点组成的无线接入点集合中随机选定两个无线接入点作为参考接入点；Randomly select two wireless access points from a wireless access point set composed of all wireless access points as reference access points;
计算两个参考接入点间的第一互信息；calculating first mutual information between two reference access points;
从无线接入点集合中获取可使第二互信息最小的无线接入点；Acquire a wireless access point that can minimize the second mutual information from the wireless access point set;
从无线接入点集合中获取可使第三互信息最小的无线接入点；Acquire a wireless access point that can minimize the third mutual information from the wireless access point set;
以此类推，直至获取到足够数量的无线接入点。And so on, until a sufficient number of wireless access points are obtained.
本实施例中，更加具体地，设定室内定位环境可用的参考点AP个数为T，选取其中S个AP的优化子集则可以将信号空间的维度从T维降到S维，因而可以减少计算量。具体地，本实施例采用互信息最小化AP选取策略，且具体步骤如下：In this embodiment, more specifically, the number of reference point APs available in the indoor positioning environment is set as T, and selecting an optimal subset of S APs among them can reduce the dimension of the signal space from the T dimension to the S dimension, so that Reduce the amount of computation. Specifically, this embodiment adopts the mutual information minimization AP selection strategy, and the specific steps are as follows:
1)对于选取的S个AP进行两两组合，按照下式计算每个组合的互信息，查找出互信息最小的组合，其对应的APm，APn作为两个初始参考点AP；1) Combining the selected S APs in pairs, calculate the mutual information of each combination according to the following formula, find out the combination with the minimum mutual information, and use its corresponding APm and APn as the two initial reference points AP;
MI(APm,APn)＝H(APm)+H(APn)-H(APm,APn) (2)MI(APm ,APn )=H(APm )+H(APn )-H(APm ,APn ) (2)
公式(2)中：MI(APm,APn)表示两个不同AP的互信息，即，第一互信息，H(APm,APn)表示两个AP的组合信息熵。In the formula (2): MI(APm , APn ) represents the mutual information of two different APs, that is, the first mutual information, and H(APm , APn ) represents the combined information entropy of the two APs.
2)按照公式(3)计算某AP与两个初始AP组合的互信息。2) Calculate the mutual information between an AP and two initial AP combinations according to formula (3).
MI(APm,APn,APi)＝H(APm,APn)+H(APi)-H(APm,APn,APi) (3)MI(APm ,APn ,APi )=H(APm ,APn )+H(APi )-H(APm ,APn ,APi ) (3)
找出能使得MI最小的AP作为最优化AP子集的第三个AP。Find the AP that can make MI the smallest as the third AP of the optimal AP subset.
3)依次按照第2)步的形式选取下一个最优的AP，依次迭代，直到选取出S个最优AP为止。需要说明的是，第R个最优的AP的选取算式如下：3) Select the next optimal AP sequentially according to the form of step 2), and iterate successively until S optimal APs are selected. It should be noted that the formula for selecting the Rth optimal AP is as follows:
MI(AP1,AP2,…,APR)＝H(AP1,AP2,…,APR-1)+H(APR)-H(APm,APn,…,APR) (4)MI(AP1 ,AP2 ,…,APR )=H(AP1 ,AP2 ,…,APR-1 )+H(APR )-H(APm ,APn ,…,APR ) ( 4)
优选地，第二匹配单元4具体用于，Preferably, the second matching unit 4 is specifically used for,
根据当前WIFI信号强度集合，以及根据预先构建的贝叶斯后验估计模型获取机器人的估计位置；Obtain the estimated position of the robot according to the current WIFI signal strength set and the pre-built Bayesian posterior estimation model;
根据估计位置，WIFI指纹地图，以及根据啮齿类动物模型中的位姿细胞网络，进行位姿信息匹配，以匹配出与当前WIFI信号强度集合具有最大指纹相似度的位姿信息。According to the estimated position, the WIFI fingerprint map, and the pose cell network in the rodent model, the pose information is matched to match the pose information with the maximum fingerprint similarity to the current WIFI signal strength set.
本实施例中，采用贝叶斯位置估计策略，且具体地，针对上述的互信息最小化AP选取策略，进一步使用贝叶斯后验估计进行组合优化，使得WIFI指纹定位算法的位置估计精度和可靠度大大提升。In this embodiment, the Bayesian position estimation strategy is adopted, and specifically, for the above-mentioned mutual information minimization AP selection strategy, Bayesian posterior estimation is further used for combined optimization, so that the position estimation accuracy of the WIFI fingerprint positioning algorithm and The reliability is greatly improved.
贝叶斯后验估计的基本原理如下式所示：The basic principle of Bayesian posterior estimation is as follows:
式中：RSSI表示多个AP在位置估计点的RSSI观测值；p(Li|RSSI)表示位置Li的在给定RSSI下的条件概率，即在观测到RSSI向量的情况下，定位点出现在Li的概率；p(RSSI|Li)表示位置Li的概率；p(Li)表示位置Li的概率，通常不考虑指纹点之间的差异，即指纹点等概率；p(RSSI)表示RSSI出现的全概率，其算式如下：In the formula: RSSI indicates the RSSI observation value of multiple APs at the location estimation point; p(Li |RSSI) indicates the conditional probability of the location Li under the given RSSI, that is, when the RSSI vector is observed, the location point The probability of appearing in Li ; p(RSSI|Li ) indicates the probability of position Li ; p(Li ) indicates the probability of position Li , usually regardless of the difference between fingerprint points, that is, the equal probability of fingerprint points; p (RSSI) indicates the total probability of RSSI occurrence, and its formula is as follows:
其中，C(RSSI1,RSSI2,…,RSSIM)表示指纹点观测到的指定RSSI向量的个数；K表示指纹点观测历元数。Among them, C(RSSI1, RSSI2,...,RSSIM) represents the number of specified RSSI vectors observed by fingerprint points; K represents the number of epochs observed by fingerprint points.
将上述全概率算式回代至贝叶斯后验估计式，从而计算出后验条件概率。采用多个指纹点的贝叶斯权重位置估计算式能够在较短时间内算出位置估计点的位置，令估计点的位置为p，则估计位置的计算公式如下：Substitute the above full probability formula back into the Bayesian posterior estimation formula to calculate the posterior conditional probability. The Bayesian weight position estimation formula using multiple fingerprint points can calculate the position of the position estimation point in a relatively short period of time. Let the position of the estimated point be p, then the calculation formula of the estimated position is as follows:
式中：(x,y)表示位置估计点的二维坐标，(xi,yi)表示第i个指纹点的坐标，ωi表征第i个指纹点的加权权重，即为贝叶斯后验条件的概率，K表示邻近点个数In the formula: (x, y) represents the two-dimensional coordinates of the position estimation point, (xi , yi ) represents the coordinates of the i-th fingerprint point, ωi represents the weighted weight of the i-th fingerprint point, which is the Bayesian The probability of the posterior condition, K represents the number of adjacent points
优选地，位姿细胞网络进行位姿信息匹配，包括：Preferably, the pose cell network performs pose information matching, including:
根据估计位置，从WIFI指纹地图中提取出与估计位置相邻的至少一个经历单元；Extracting at least one experience unit adjacent to the estimated position from the WIFI fingerprint map according to the estimated position;
计算每个经历单元的WIFI信号强度与当前WIFI信号强度间的欧氏距离；Calculate the Euclidean distance between the WIFI signal strength of each experience unit and the current WIFI signal strength;
获取最大欧式距离对应的经历单元所指向的位姿信息，并将获取到的位姿信息确定为机器人的当前位姿信息。Obtain the pose information pointed to by the experienced unit corresponding to the maximum Euclidean distance, and determine the obtained pose information as the current pose information of the robot.
进一步优选地，欧式距离的计算公式为：Further preferably, the formula for calculating the Euclidean distance is:
其中，(xpc,ypc,θpc)为经历单元对应的位姿细胞坐标；(xi,yi,θi)为与当前位置对应的位姿细胞坐标；ra为(x,y)平面的区域常数，θa为θ维上的区域常数。Among them, (xpc ,ypc ,θpc ) are the coordinates of the pose cell corresponding to the experienced unit; (xi ,yi ,θi ) are the coordinates of the pose cell corresponding to the current position; ra is (x,y ) plane area constant, θa is the area constant on the θ dimension.
本实施例中，每个经历单元都具有一个活性水平，活性水平由位姿感知细胞和WIFI指纹中能量峰与每一个经历单元之间接近程度所决定。每一个经历在位姿感知细胞和WIFI指纹中有一个相关的活性区域。当能量峰处于这些活性区域时，该激励立刻被激活，这些区域在位姿感知细胞内部是连续的，而在WIFI指纹中的相关区域却是非连续的。每个经历ei由经历活性水平Ei，WIFI信号强度Ri所决定。其中，ei＝{Ei,Ri}，In this embodiment, each experience unit has an activity level, and the activity level is determined by the proximity between the pose sensing cells and the energy peak in the WIFI fingerprint and each experience unit. Each experience has an associated active area in the pose sensing cells and WIFI fingerprints. When the energy peak is in these active areas, the excitation is activated immediately, and these areas are continuous inside the pose sensing cells, but the relevant areas in the WIFI fingerprint are discontinuous. Each experience ei is determined by the experience activity level Ei and the WIFI signal strength Ri . Among them, ei = {Ei , Ri },
一个经历单元的能级Exyθ和第i个经历单元的总能级水平Ei由公式(9)和公式(10)计算可得。The energy level Exyθ of an experienced unit and the total energy level Ei of the i-th experienced unit can be calculated by formula (9) and formula (10).
其中，xpcypc和θpc为最大活性姿态细胞的坐标；xi、yi、θi为与该经历相关的位姿感知细胞的坐标；ra为(x,y)平面的区域常数；θa为θ维上的区域常数。Rcurr为当前WIFI信号强度；Ri为与经历i相关的WIFI信号强度。Among them, xpc ypc and θpc are the coordinates of the most active attitude cell; xi , yi , θi are the coordinates of the pose sensing cells related to the experience; ra is the area constant of the (x,y) plane ; θa is the area constant on the θ dimension. Rcurr is the current WIFI signal strength; Ri is the WIFI signal strength related to experience i.
需要说明的是，本实施例中，如图3所示地，利用WIFI指纹匹配过程关系示意图。将无线信号网络WIFI作为一种传感器用到啮齿类动物模型当中，其定位模型有三个主要部分组成，分别为WIFI指纹，位姿细胞网络和经历图。WIFI指纹获取环境的WIFI信号强度，被称为WIFI信号强度模板。WIFI指纹信息用来辨识熟悉的环境。当新输入的WIFI信号强度信息与已存在的WIFI信号强度模板匹配时，位姿细胞网络的活性因子被激活，二者结合能够很大程度上阻止错误匹配的发生，产生更为准确的经历图。It should be noted that, in this embodiment, as shown in FIG. 3 , it is a schematic diagram of a process relationship using WIFI fingerprint matching. The wireless signal network WIFI is used as a sensor in the rodent model. The positioning model consists of three main parts, namely WIFI fingerprint, pose cell network and experience map. The WIFI fingerprint acquires the WIFI signal strength of the environment, which is called a WIFI signal strength template. WIFI fingerprint information is used to identify familiar environments. When the newly input WIFI signal strength information matches the existing WIFI signal strength template, the activity factor of the pose cell network is activated. The combination of the two can largely prevent the occurrence of wrong matching and generate a more accurate experience map. .
尽管本发明已进行了一定程度的描述，明显地，在不脱离本发明的精神和范围的条件下，可进行各个条件的适当变化。可以理解，本发明不限于所述实施方案，而归于权利要求的范围，其包括所述每个因素的等同替换。While the invention has been described to a certain extent, it will be obvious that various changes may be made in various conditions without departing from the spirit and scope of the invention. It is to be understood that the invention is not limited to the described embodiments, but rather falls within the scope of the claims, which include equivalents to each of the elements described.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2017113208868 | 2017-12-12 | ||
| CN201711320886 | 2017-12-12 | ||
| CN2018101033693 | 2018-02-01 | ||
| CN201810103369 | 2018-02-01 |
| Publication Number | Publication Date |
|---|---|
| CN108680175Atrue CN108680175A (en) | 2018-10-19 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810265940.1APendingCN108680175A (en) | 2017-12-12 | 2018-03-28 | Synchronous superposition method and device based on rodent models |
| Country | Link |
|---|---|
| CN (1) | CN108680175A (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112230243A (en)* | 2020-10-28 | 2021-01-15 | 西南科技大学 | Indoor map construction method for mobile robot |
| CN112325883A (en)* | 2020-10-19 | 2021-02-05 | 湖南大学 | Indoor positioning method for mobile robot with WiFi and visual multi-source integration |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104244409A (en)* | 2014-09-30 | 2014-12-24 | 智慧城市信息技术有限公司 | Server and method and system for implementing positioning by utilizing wireless signals |
| CN104507050A (en)* | 2014-11-24 | 2015-04-08 | 北京航空航天大学 | A method for probability type fingerprint matching in WiFi (Wireless Fidelity) indoor positioning |
| CN104640201A (en)* | 2014-12-31 | 2015-05-20 | 智慧城市信息技术有限公司 | Positioning method and device based on WIFI (wireless fidelity) fingerprint technology |
| CN105378809A (en)* | 2013-03-14 | 2016-03-02 | 谷歌公司 | Apparatus, method and associated information processing for security of home provided with smart sensors |
| CN106793075A (en)* | 2016-12-21 | 2017-05-31 | 武汉大学 | A kind of WiFi indoor orientation methods based on domain cluster |
| CN106814737A (en)* | 2017-01-20 | 2017-06-09 | 安徽工程大学 | A kind of SLAM methods based on rodent models and RTAB Map closed loop detection algorithms |
| CN107409051A (en)* | 2015-03-31 | 2017-11-28 | 深圳市大疆创新科技有限公司 | Authentication system and method for generating flight controls |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105378809A (en)* | 2013-03-14 | 2016-03-02 | 谷歌公司 | Apparatus, method and associated information processing for security of home provided with smart sensors |
| CN104244409A (en)* | 2014-09-30 | 2014-12-24 | 智慧城市信息技术有限公司 | Server and method and system for implementing positioning by utilizing wireless signals |
| CN104507050A (en)* | 2014-11-24 | 2015-04-08 | 北京航空航天大学 | A method for probability type fingerprint matching in WiFi (Wireless Fidelity) indoor positioning |
| CN104640201A (en)* | 2014-12-31 | 2015-05-20 | 智慧城市信息技术有限公司 | Positioning method and device based on WIFI (wireless fidelity) fingerprint technology |
| CN107409051A (en)* | 2015-03-31 | 2017-11-28 | 深圳市大疆创新科技有限公司 | Authentication system and method for generating flight controls |
| CN106793075A (en)* | 2016-12-21 | 2017-05-31 | 武汉大学 | A kind of WiFi indoor orientation methods based on domain cluster |
| CN106814737A (en)* | 2017-01-20 | 2017-06-09 | 安徽工程大学 | A kind of SLAM methods based on rodent models and RTAB Map closed loop detection algorithms |
| Title |
|---|
| RAFAEL BERKVENS1, ET AL: "Biologically Inspired SLAM Using Wi-Fi", 《2014 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS 2014)》* |
| RAFAEL BERKVENS1, ET AL: "Biologically Inspired SLAM Using Wi-Fi", 《2014 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS》* |
| 高晓颖: "《源于自然的机器人导航 基于啮齿类动物模型的同步定位地图构建和路径规划》", 31 January 2016, 国防工业出版社* |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112325883A (en)* | 2020-10-19 | 2021-02-05 | 湖南大学 | Indoor positioning method for mobile robot with WiFi and visual multi-source integration |
| CN112230243A (en)* | 2020-10-28 | 2021-01-15 | 西南科技大学 | Indoor map construction method for mobile robot |
| CN112230243B (en)* | 2020-10-28 | 2022-04-08 | 西南科技大学 | Indoor map construction method for mobile robot |
| Publication | Publication Date | Title |
|---|---|---|
| CN110856112B (en) | Crowd-sourcing perception multi-source information fusion indoor positioning method and system | |
| CN109682382B (en) | Global fusion localization method based on adaptive Monte Carlo and feature matching | |
| CN110118549B (en) | Multi-source information fusion positioning method and device | |
| US20180245927A1 (en) | Navigate, track, and position mobile devices in gps-denied or gps-inaccurate areas with automatic map generation | |
| CN110375739B (en) | A mobile terminal vision fusion positioning method, system and electronic device | |
| CN105636201B (en) | Indoor orientation method based on sparse signal fingerprint database | |
| WO2019061949A1 (en) | Motion-behavior-assisted indoor fusion positioning method and apparatus and storage medium | |
| CN110118556A (en) | A kind of robot localization method and device based on covariance mixing together SLAM | |
| CN103889049A (en) | Wireless signal indoor positioning system and method based on inertia measurement element assistance | |
| CN108680177B (en) | Synchronous positioning and map construction method and device based on rodent model | |
| CN104881029B (en) | Mobile Robotics Navigation method based on a point RANSAC and FAST algorithms | |
| CN107270889B (en) | An indoor positioning method and positioning system based on geomagnetic atlas | |
| CN106767828A (en) | A kind of mobile phone indoor positioning solution | |
| CN115267796A (en) | Positioning method, positioning device, robot and storage medium | |
| CN113447014A (en) | Indoor mobile robot, mapping method, positioning method, and mapping positioning device | |
| CN108919177A (en) | A Location Map Construction Method Based on Virtual Source Estimation and Trajectory Correction | |
| KR102616201B1 (en) | A positioning method based on artificial intelligence neural network based on sensor map images of city data and an apparatus using it | |
| Feng et al. | Visual Map Construction Using RGB‐D Sensors for Image‐Based Localization in Indoor Environments | |
| CN105898711B (en) | A kind of localization method and device based on earth magnetism reference line | |
| CN108712725B (en) | A SLAM method based on rodent model and WIFI fingerprint | |
| CN108680175A (en) | Synchronous superposition method and device based on rodent models | |
| CN120086416A (en) | WiFi positioning method and related device based on multimodal knowledge base | |
| CN116563352A (en) | Single-line laser radar loop detection method and system integrating depth visual information | |
| CN110348333A (en) | Object detecting method, device, storage medium and electronic equipment | |
| CN111836188B (en) | An online co-location and system based on Wi-Fi RSS |
| 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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication | Application publication date:20181019 |