Path planning method based on 4D risk occupation under vehicle and road cloud cooperation framework

By performing edge cloud computing and local path search under the framework of vehicle-road cloud collaboration, filtering and optimizing path planning, the problem of high complexity of path planning algorithms and perceived blind spots is solved, more efficient and accurate path planning is achieved, and driving is improved Safety and adaptability.

CN118980386APending Publication Date: 2024-11-19TSINGHUA UNIVERSITY

Patent Information

Application Number
CN202411130353.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-16
Publication Date
2024-11-19

AI Technical Summary

Technical Problem

In the prior art, the calculation complexity of the path planning algorithm is high, resulting in slow operation and inability to meet the real-time requirements. The occlusion and blind spot problems of bicycle perception limit the comprehensiveness and accuracy of path planning, increasing uncertainty during driving. Sex and risk.

Method used

Under the framework of vehicle-road cloud collaboration, by obtaining the prior road information of the edge cloud computing platform and the perception result data of the road-end perception unit, risk value evaluation is carried out, and the feasible path nodes below the preset risk threshold are selected, and local Path search strategies and heuristic functions calculate the value of the generation, select the optimal path, and generate the final path through Gaussian filtering optimization.

Benefits of technology

Improve computing efficiency, alleviate memory pressure, expand the perception range, improve adaptability in different traffic environments, and reduce uncertainty and risks during driving.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118980386A_ABST
    Figure CN118980386A_ABST
Patent Text Reader

Abstract

The invention relates to a 4D risk occupation-based path planning method under a vehicle-road cloud collaborative framework, and the method comprises the steps: obtaining a potential path node set according to prior road information pre-stored by an edge cloud computing platform and sensing result data of a road end sensing unit in combination with an instruction sent by an ICV; performing risk value assessment on the potential path node set, and retaining initial drivable path nodes lower than a risk threshold value to generate a dynamic collision-free node set; based on the dynamic collision-free node set, a local path search strategy is used for obtaining the node range of each batch in front of the intelligent network connection vehicle, the cost of each node is calculated in the range, the node with the minimum cost is selected as the path node of the batch, the next batch is processed according to the path node, and a final local path is obtained. Therefore, the problems that the uncertainties and risks in the driving process are possibly increased due to the fact that the calculation complexity is high, a blind area is sensed by a bicycle and the adaptability to a specific scene is insufficient in the prior art are solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to a path planning method based on 4D risk occupancy in a vehicle-road-cloud collaborative framework. Background Art

[0002] In recent years, with the rapid development of intelligent transportation and autonomous driving, path planning technology has been widely studied and applied to ensure safe and efficient driving of vehicles.

[0003] In related technologies, the time dimension can be added to the two-dimensional space model to create a three-dimensional space-time risk field to guide trajectory planning. The risk field model can be improved, and the A* algorithm can be optimized through adaptive strategies and loss functions. The kinematic characteristics of the vehicle can be considered to provide effective trajectory planning for intelligent vehicles in complex traffic environments.

[0004] However, in related technologies, due to the high computational complexity, the algorithm may run slowly and fail to meet real-time requirements, affecting the vehicle's ability to respond quickly. Due to the occlusion and blind spot problems in single-vehicle perception, the comprehensiveness and accuracy of path planning may be limited, increasing uncertainty and potential risks during driving. It may also lack sufficient flexibility and adaptability, limiting the breadth and reliability of application, and urgently needs to be improved. Summary of the invention

[0005] The present application provides a path planning method based on 4D risk occupancy under the vehicle-road-cloud collaborative framework to alleviate the problems of related technologies such as high computational complexity, occlusion blind spots of single-vehicle perception, and insufficient adaptability to specific scenarios, which may increase the uncertainty and risk of the driving process.

[0006] In a first aspect, an embodiment of the present application provides a path planning method based on 4D risk occupancy under a vehicle-road-cloud collaborative framework, comprising the following steps: obtaining a priori road information pre-stored by an edge cloud computing platform and perception result data of multiple road-end perception units in a target area, so that the edge cloud computing platform obtains a potential path node set for the target intelligent connected vehicle according to the priori road information, the perception result data and the instruction information sent by the target intelligent connected vehicle; performing risk value evaluation calculation on each potential path node in the potential path node set to obtain a risk value evaluation result of each potential path node, and performing risk value evaluation calculation on each potential path node according to the risk value evaluation result. The initial drivable path nodes below the preset risk threshold are retained to obtain a dynamic collision-free node set; based on the dynamic collision-free node set, the drivable nodes of each batch are limited in range by using a preset local path search strategy, and the cost of each drivable node is calculated by using a preset heuristic function within the limited range to obtain a cost value of each drivable node, so as to select each drivable node with the lowest cost value within the limited batch range as the optimal drivable node of the corresponding batch according to the cost value, and the final local path for the target intelligent connected vehicle is obtained based on the drivable nodes of each batch and the preset Gaussian filtering strategy.

[0007] Optionally, in one embodiment of the present application, based on the dynamic collision-free node set, a preset local path search strategy is used to limit the range of the drivable nodes of each batch, and a preset heuristic function is used to calculate the cost of each drivable node within the limited range to obtain a cost value of each drivable node, so as to select each drivable node with the lowest cost value within the limited batch range as the optimal drivable node of the corresponding batch according to the cost value, and based on the drivable nodes of each batch and a preset Gaussian filtering strategy, a final local path facing the target intelligent connected vehicle is obtained, including: obtaining the vehicle position of the target intelligent connected vehicle according to the instruction information sent by the target intelligent connected vehicle, so as to determine the initial neighboring nodes of the target intelligent connected vehicle through the vehicle position, and based on the initial neighboring nodes, determine the first batch of neighboring nodes in combination with the preset movement rules and the dynamic collision-free node set; obtaining The perception result data of each frame of the edge cloud computing platform and the multiple road-side perception units are obtained, and the risk value of each drivable node is calculated according to the perception result data of each frame, and the risk value of each potential node is screened and filtered to retain each drivable node below the preset risk threshold, and a new dynamic collision-free node set is constructed; the search range of each drivable node in each batch is limited by the local path search strategy, and within the limited range, the cost value of each drivable node is calculated by the heuristic function, and the drivable node with the smallest cost value is selected as the optimal node of the corresponding batch; the optimal nodes of each batch are added to the preset closed set in turn to obtain the optimal node closed set, and the Gaussian filtering optimization operation is performed on the optimal node closed set to obtain the final optimal local path node set, and based on the final optimal local path node set, the final local path for the target intelligent connected vehicle is generated.

[0008] Optionally, in one embodiment of the present application, the cost of each drivable node is calculated using a preset heuristic function within a limited range to obtain the cost value of each drivable node, including: constructing the heuristic function based on the Euclidean distance cost from the current drivable node to the end point on the Cartesian coordinate system, the Euclidean distance cost between the target intelligent connected vehicle and the end point, and the road risk cost; based on the dynamic collision-free node set and the heuristic function, calculating the cost value of each drivable node in each batch within the limited range.

[0009] Optionally, in one embodiment of the present application, the expression of the heuristic function is:

[0010]

[0011] Among them, risk represents the cost value of each drivable node; dest(x, y) represents the coordinates of the destination; sample(x, y) and ICV(x, y) represent the coordinates of the sampling point and the target intelligent connected vehicle respectively.

[0012] The second aspect of the present application provides a path planning device based on 4D risk occupancy under the vehicle-road-cloud collaborative framework, including: an acquisition module, used to acquire the prior road information pre-stored by the edge cloud computing platform and the perception result data of multiple road-end perception units in the target area, so that the edge cloud computing platform obtains a potential path node set for the target intelligent networked vehicle according to the prior road information, the perception result data and the instruction information sent by the target intelligent networked vehicle; a risk assessment module, used to perform risk value assessment calculation on each potential path node in the potential path node set, obtain the risk value assessment result of each potential path node, and The result retains the initial drivable path nodes that are lower than the preset risk threshold to obtain a dynamic collision-free node set; a path planning module is used to limit the range of drivable nodes in each batch based on the dynamic collision-free node set by using a preset local path search strategy, and calculate the cost of each drivable node within the limited range by using a preset heuristic function to obtain a cost value of each drivable node, so as to select each drivable node with the lowest cost value within the limited batch range as the optimal drivable node of the corresponding batch according to the cost value, and obtain the final local path for the target intelligent connected vehicle based on the drivable nodes of each batch and a preset Gaussian filtering strategy.

[0013] Optionally, in one embodiment of the present application, the path planning module includes: a determination unit, used to obtain the vehicle position of the target intelligent connected vehicle according to the instruction information sent by the target intelligent connected vehicle, so as to determine the initial neighboring nodes of the target intelligent connected vehicle through the vehicle position, and determine the first batch of neighboring nodes based on the initial neighboring nodes, while combining the preset movement rules and the dynamic collision-free node set; a retention unit, used to obtain the perception result data of each frame of the edge cloud computing platform and the multiple road-side perception units, and calculate the risk value of each drivable node based on the perception result data of each frame, and screen and filter the risk value of each potential node to retain those below the preset risk threshold. each drivable node, and construct a new dynamic collision-free node set; a first calculation unit, used to use the local path search strategy to limit the search range of each drivable node in each batch, and within the limited range, calculate the cost value of each drivable node through the heuristic function, and select the drivable node with the smallest cost value as the optimal node of the corresponding batch; a filtering unit, used to add the optimal nodes of each batch to the preset closed set in turn to obtain the optimal node closed set, and perform Gaussian filtering optimization operation on the optimal node closed set to obtain the final optimal local path node set, and based on the final optimal local path node set, generate the final local path for the target intelligent connected vehicle.

[0014] Optionally, in one embodiment of the present application, the path planning module also includes: a construction unit, used to construct the heuristic function based on the Euclidean distance cost from the current drivable node to the end point in the Cartesian coordinate system, the Euclidean distance cost between the target intelligent connected vehicle and the end point, and the road risk cost; a second calculation unit, used to calculate the cost value of each drivable node in each batch within the limited range based on the dynamic collision-free node set and the heuristic function.

[0015] Optionally, in one embodiment of the present application, the mathematical expression of the heuristic function is:

[0016]

[0017] Wherein, risk represents the cost value of each drivable node; dest(x, y) represents the coordinates of the destination; sample(x, y) and ICV(x, y) represent the coordinates of the sampling point and the target intelligent connected vehicle respectively.

[0018] The third aspect of the present application provides a vehicle, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement a path planning method based on 4D risk occupancy under the vehicle-road-cloud collaborative framework as described in the above embodiment.

[0019] The fourth aspect embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the program is executed by a processor, it implements the above-mentioned path planning method based on 4D risk occupancy under the vehicle-road-cloud collaborative framework.

[0020] The fifth aspect embodiment of the present application provides a computer program product, including a computer program, which, when executed, is used to implement the above-mentioned path planning method based on 4D risk occupancy under the vehicle-road-cloud collaborative framework.

[0021] The embodiment of the present application can obtain the prior road information pre-stored by the edge cloud computing platform and the perception result data of multiple road-side perception units in the target area, so that the edge cloud computing platform can obtain a potential path node set for the target intelligent connected vehicle according to the prior road information, the perception result data and the instruction information sent by the target intelligent connected vehicle; perform risk value evaluation calculation on each potential path node in the potential path node set to obtain the risk value evaluation result of each potential path node, and retain the initial drivable path nodes below the preset risk threshold according to the risk value evaluation result to obtain a dynamic collision-free node set; based on the dynamic collision-free node set, use a preset local path search strategy to search for each potential path node in the potential path node set; The range of drivable nodes in batches is limited, and the preset heuristic function is used to calculate the cost of each drivable node within the limited range to obtain the cost value of each drivable node, so as to select each drivable node with the lowest cost value in the limited batch range as the optimal drivable node of the corresponding batch according to the cost value, and the final local path for the target intelligent connected vehicle is obtained based on the drivable nodes of each batch and the preset Gaussian filtering strategy, so as to effectively improve the computing efficiency, alleviate the memory pressure brought by large-scale data, expand the perception range, and improve the adaptability in different traffic environments by using the risk occupancy assessment algorithm and the local path planning algorithm as well as the vehicle-road-cloud collaborative architecture. As a result, the problems of high computational complexity, blind spot occlusion of single-vehicle perception and insufficient adaptability to specific scenarios in related technologies, which may increase the uncertainty and risk of the driving process, are alleviated.

[0022] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0024] Figure 1 A flowchart of a path planning method based on 4D risk occupancy in a vehicle-road-cloud collaborative framework provided according to an embodiment of the present application;

[0025] Figure 2 A schematic diagram of 4D risk occupation according to an embodiment of the present application;

[0026] Figure 3 A schematic diagram of the principle of a local path planning method according to an embodiment of the present application;

[0027] Figure 4 A schematic diagram of a vehicle-road-cloud collaboration framework according to an embodiment of the present application;

[0028] Figure 5 A schematic diagram of the structure of a path planning device based on 4D risk occupancy in a vehicle-road-cloud collaborative framework provided according to an embodiment of the present application;

[0029] Figure 6 It is a schematic diagram of the structure of a vehicle provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0030] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0031] The following describes the path planning method based on 4D risk occupancy under the vehicle-road-cloud collaborative framework of the embodiment of the present application with reference to the accompanying drawings. In view of the related technologies mentioned in the above background technology center, due to the high computational complexity, the occlusion blind spot problem of single-vehicle perception and the lack of adaptability to specific scenarios, the real-time, comprehensive and reliability of path planning are limited, which may increase the uncertainty and risk in the driving process. The present application provides a path planning method based on 4D risk occupancy under the vehicle-road-cloud collaborative framework. In this method, by obtaining the prior road information pre-stored by the edge cloud computing platform and the perception result data of multiple road-end perception units in the target area, the edge cloud computing platform obtains a potential path node set for the target intelligent connected vehicle based on the prior road information, the perception result data and the instruction information sent by the target intelligent connected vehicle; a risk value evaluation calculation is performed on each potential path node in the potential path node set to obtain the risk value evaluation result of each potential path node, and the risk value evaluation result is retained according to the risk value evaluation result. The initial drivable path nodes below the preset risk threshold are used to obtain a dynamic collision-free node set; based on the dynamic collision-free node set, the preset local path search strategy is used to limit the range of drivable nodes in each batch, and the preset heuristic function is used to calculate the cost of each drivable node within the limited range to obtain the cost value of each drivable node, so as to select each drivable node with the lowest cost value within the limited batch range as the optimal drivable node of the corresponding batch according to the cost value, and the final local path for the target intelligent connected vehicle is obtained based on the drivable nodes of each batch and the preset Gaussian filtering strategy. The algorithm is deployed on the edge cloud computing platform, thereby using the risk occupancy assessment algorithm and the local path planning algorithm as well as the vehicle-road-cloud collaborative architecture to effectively improve the computing efficiency, relieve the memory and computing pressure of the vehicle-side platform, expand the perception range, and improve the adaptability in different traffic environments. As a result, the problems of high computational complexity, blind spot occlusion of single-vehicle perception, and insufficient adaptability to specific scenarios in related technologies, which may increase the uncertainty and risk of the driving process, are alleviated.

[0032] Specifically, Figure 1 A flow chart of a path planning method based on 4D risk occupancy in a vehicle-road-cloud collaborative framework provided in an embodiment of the present application.

[0033] like Figure 1 As shown, the path planning method based on 4D risk occupancy under the vehicle-road-cloud collaborative framework includes the following steps:

[0034] In step S101, the prior road information pre-stored by the edge cloud computing platform and the perception result data of multiple road-end perception units in the target area are obtained, so that the edge cloud computing platform obtains a potential path node set for the target intelligent connected vehicle based on the prior road information, the perception result data and the instruction information sent by the target intelligent connected vehicle.

[0035] Those skilled in the art should understand that the embodiments of the present application can perform 4D risk occupancy assessment on the road surface space of traffic roads by combining the ideas of the occupancy grid map field and the traditional driving risk field modeling field, wherein the 4D risk occupancy includes 2-dimensional road plane space occupancy (x, y), 1-dimensional risk value (z) and 1-dimensional time dimension occupancy prediction (2-3s), such as Figure 2 As shown, it is possible to judge which spaces on the road are occupied by risks and which spaces are drivable because the risks are below the threshold under the unified perception form of the road space environment.

[0036] Specifically, in an embodiment of the present application, the edge cloud computing platform obtains the set of future potential path nodes required for the ICV to enter the intersection through the perception data obtained by multiple road-side perception units, the data stored by the edge cloud computing platform based on the prior road information, and the instructions issued by the ICV. The embodiment of the present application can also convert dynamic and static factors into geometric objects based on the prior road information pre-stored by the platform and the perception result data from multiple road-side perception units in the area, and obtain the risk value of each node through the interaction between the node and the geometric object, and store it in each corresponding node.

[0037] Understandably, the vehicle-road-cloud collaborative framework emphasizes the scalability of the system and user-targeted dynamic services. With the emergence of new technologies and the evolution of traffic needs, the framework can flexibly integrate new functions and services to adapt to the ever-changing traffic environment. At the same time, the framework supports the provision of dynamic traffic road perception range and limited potential path nodes according to the specific needs of ICV, alleviating problems such as computational timeliness and memory resource pressure; in addition, the framework can customize safe, beyond-visual-range driving paths based on the driving needs and related information of ICV; in short, the vehicle-road-cloud collaborative framework represents the future direction of the development of intelligent transportation systems, and through the deep integration and collaborative work of vehicles, roads, and clouds, it lays a solid foundation for achieving safe, efficient, and intelligent transportation goals.

[0038] In step S102, a risk value assessment is performed on each potential path node in the potential path node set to obtain a risk value assessment result for each potential path node, and initial drivable path nodes below a preset risk threshold are retained based on the risk value assessment result to obtain a dynamic collision-free node set.

[0039] Those skilled in the art should understand that the edge cloud computing platform is capable of processing and analyzing any computing resource or system of a dynamic collision-free node set; in addition, in order to generate a dynamic collision-free node set, the rules followed by the embodiments of the present application during processing and updating include but are not limited to collision avoidance, connectivity, dynamic constraints, etc.

[0040] It will be understood that the preset risk threshold refers to a pre-set risk value for assessing a potential driving path, which can be specifically set by those skilled in the art based on factors such as road conditions, traffic density, weather conditions, vehicle speed, and the distance and dynamics of potential obstacles, and is not specifically limited here.

[0041] During the actual execution process, the embodiment of the present application can retrieve the road point sets corresponding to different driving behavior instructions in advance based on the current position of the ICV, action instructions and the road area where it is located, and based on the areas that may be reached in the future when the driving action is completed under legal conditions. Then, a risk threshold is set, and the road point sets that are smaller than the risk threshold are filled into the dynamic collision-free node set, which is updated again in each frame to support dynamic planning.

[0042] Specifically, in the embodiment of the present application, when planning the driving path of a vehicle from point A to point B, the sensors of multiple road-side sensing units can be used to obtain dynamic obstacles on the road (such as other vehicles and pedestrians), and the road layout, static obstacles (such as buildings and trees), and the types and locations of road signs and traffic light poles, etc. can be obtained through prior information. Then, risk assessment can be performed on each possible driving path node. It should be noted that in the embodiment of the present application, a node represents a circular road surface space with a radius of 1-2m, and the risk value needs to be normalized to the range of 0-1. The risk threshold can generally be set to 0.15. If it is higher than the risk threshold, it means that the circular road surface space is not feasible and other adjacent nodes of the same batch need to be found.

[0043] In the specific implementation process, before the ICV is ready to enter the intersection, the edge cloud platform can understand its driving intention through the command information sent by the ICV, that is, obtain the ICV's intention to turn left, turn right or go straight, and the embodiment of the present application can use the local path planning algorithm to find the best path according to the potential path node set of the corresponding intention, and the risk value assessment of the node is to assess which spaces in the drivable area are occupied by risks and which spaces are below the risk threshold (that is, drivable space nodes).

[0044] The embodiment of the present application screens out the initial drivable path nodes that are below a preset risk threshold, which helps to simplify the complexity of the road environment into an intuitive risk assessment, better understand the surrounding environment, and effectively improve computing efficiency and avoid potential dangers by excluding high-risk driving paths.

[0045] In step S103, based on the dynamic collision-free node set, the preset local path search strategy is used to limit the range of the drivable nodes of each batch, and the preset heuristic function is used to calculate the cost of each drivable node within the limited range to obtain the cost value of each drivable node, so as to select each drivable node with the lowest cost value within the limited batch range as the optimal drivable node of the corresponding batch according to the cost value, and the final local path for the target intelligent connected vehicle is obtained based on the drivable nodes of each batch and the preset Gaussian filtering strategy.

[0046] In the actual implementation process, the embodiment of the present application can take advantage of the edge cloud computing platform and combine it with the vehicle-road-cloud collaborative framework to adapt to the ever-changing traffic environment. Among them, the embodiment of the present application can consider the future driving needs of ICVs, combine road prior information and real-time dynamic data, and process and update the dynamic collision-free node set.

[0047] The embodiment of the present application utilizes an edge cloud computing platform to process and update a dynamic collision-free node set and performs data processing at the edge of the network, which can effectively reduce the computational complexity of the algorithm and reduce data transmission delays, so that the path planning algorithm can respond quickly and meet real-time requirements.

[0048] Optionally, in one embodiment of the present application, based on a dynamic collision-free node set, a preset local path search strategy is used to limit the range of the drivable nodes of each batch, and a preset heuristic function is used to calculate the cost of each drivable node within the limited range to obtain a cost value for each drivable node, so as to select each drivable node with the lowest cost value within the limited batch range as the optimal drivable node of the corresponding batch according to the cost value, and based on the drivable nodes of each batch and a preset Gaussian filtering strategy, a final local path for the target intelligent connected vehicle is obtained, including: obtaining the vehicle position of the target intelligent connected vehicle according to the instruction information sent by the target intelligent connected vehicle, so as to determine the initial neighboring nodes of the target intelligent connected vehicle through the vehicle position, and determining the neighboring nodes of the first batch based on the initial neighboring nodes, while combining the preset movement rules and the dynamic collision-free node set. ; Obtain the perception result data of each frame of the edge cloud computing platform and multiple road-side perception units, and calculate the risk value of each drivable node based on the perception result data of each frame, and filter the risk value of each potential node to retain each drivable node below the preset risk threshold, and build a new dynamic collision-free node set; Use the local path search strategy to limit the search range of each drivable node in each batch, and within the limited range, calculate the cost value of each drivable node through the heuristic function, and select the drivable node with the smallest cost value as the optimal node of the corresponding batch; Add the optimal node of each batch to the preset closed set in turn to obtain the optimal node closed set, and perform Gaussian filtering optimization operation on the optimal node closed set to obtain the final optimal local path node set, and based on the final optimal local path node set, generate the final local path for the target intelligent connected vehicle.

[0049] In the specific implementation process, the embodiments of the present application can retrieve a set of potential path nodes pre-constructed according to prior information based on the instruction information sent from the ICV to the edge cloud platform, and select the initial neighboring nodes according to the position of the ICV itself in the instruction, and based on the initial neighboring nodes, determine the first batch of neighboring nodes according to preset movement rules and dynamic collision-free node sets.

[0050] Furthermore, the embodiments of the present application can obtain real-time perception result data from the edge cloud platform and multiple road-side perception units, calculate the risk value of each node, and store it in each node data, and each time the edge cloud obtains a frame of perception data, the risk value of each node is recalculated.

[0051] It can be understood that the above-mentioned local path search strategy refers to a pre-set method for finding potential driving paths in a given dynamic collision-free node set while considering distance and risk factors.

[0052] Through iterative optimization, the embodiments of the present application help to ensure that all points on the path are located in a safe area, effectively predict current and future road risk occupancy conditions, identify potential dangers in advance and plan safe paths, significantly improve driving safety, and the real-time update capability of the dynamic collision-free node set ensures the dynamic adaptability of path planning, can respond quickly to immediate changes in the traffic environment, adapt to different traffic scenarios and working conditions, and improve applicability and robustness in diverse traffic environments.

[0053] It should be noted that the preset movement rules refer to the movement rules for the vehicle to move from the current node to the next node that are pre-set based on the future behavior of the ICV and traffic regulations.

[0054] Specifically, the embodiment of the present application can determine the end point based on future driving actions and obtain the dynamic collision-free node set at the current moment. Furthermore, all possible road nodes can be determined based on the ICV position and the road travel direction, and the node with the minimum cost value is selected from each batch of neighboring nodes. Based on this node, the next batch of neighboring nodes can be searched according to the rules, and the process is repeated until the end point is reached or the last node under the premise that the end point cannot be reached, and the search is ended and the path result is returned. The embodiment of the present application can sequentially add the node with the minimum cost value in each batch of neighboring nodes to the closed set, and perform Gaussian filtering optimization on the closed set to obtain the final optimal local path node set.

[0055] In the actual implementation process, the embodiment of the present application can use sensors of multiple road-side perception units to obtain dynamic obstacles on the road (such as other vehicles and pedestrians), and obtain road layout, static obstacles (such as buildings and trees) and road signs and traffic light poles through prior information. Type, location and other data, and then use the edge cloud platform to process and update the dynamic collision-free node set, evaluate all initial drivable path nodes, and if the risk assessment result of a node is found to be equal to or higher than 15%, the node will be marked as high risk and removed from the path planning. For example, if an emergency vehicle is passing ahead or there is a traffic jam, the node risk in these areas may rise, exceeding the preset 15% risk threshold, and thus be removed. At the same time, check whether there are new low-risk nodes that can be added to the path planning. For example, if a road that was originally closed due to construction is now reopened and its risk is assessed to be less than 15%, then these nodes can be added to the path planning, become part of the alternative path, and be integrated into a dynamic collision-free node set.

[0056] It should be noted that the edge cloud platform can perform risk occupancy assessment on the road space represented by each node in the drivable area (i.e., the potential path node set) retrieved according to the ICV driving intention. If the risk value is greater than the preset risk threshold of 0.15, it indicates that the local road space is not drivable, and the path algorithm can search for the optimal local path in the potential space.

[0057] For example, in planning the driving path of a vehicle from point A to point B, the embodiment of the present application originally planned to turn left at an intersection. However, through the continuous acquisition of environmental information and the use of the edge cloud platform to process and update the dynamic collision-free node set, if a traffic accident is found at the intersection ahead, causing the risk value of all path nodes in the left-turn potential point set area to exceed 0.15, that is, the left-turn area is occupied by risk, the embodiment of the present application can adjust the ICV navigation route and the corresponding ICV driving intention, remove the left-turn potential path node set, and re-retrieve a new potential path node set; if the left-turn potential path node set is partially occupied by risk, the local path planning algorithm continues to search for the optimal local path in the potential path node set.

[0058] The embodiment of the present application combines prior information of the road with real-time dynamic data to re-evaluate the risk of each initial drivable path node, which helps to avoid potential risks, optimize the driving path, and optimize the use of computing resources, reduce dependence on storage resources, and improve the operating efficiency of the algorithm.

[0059] In addition, the embodiment of the present application can number each row of the road point set in advance based on the prior knowledge of the road spatial position, and then determine the movement rules according to the future behavior of the ICV and traffic regulations to determine the next batch of neighboring nodes. Figure 3 As shown, for example, when the ICV needs to enter the intersection and turn left in the future, each batch of nodes can only select the straight direction and one node to the left in the next row of nodes, gradually guiding the ICV to move to the left as early as possible before entering the intersection and prepare to enter the rightmost lane to complete the right turn, and the trajectory generated by the RRF-based path planning algorithm is a rough trajectory connected by short line segments. In addition, in order to further improve the smoothness and practicality of the planned trajectory, the algorithm uses Gaussian trajectory smoothing.

[0060] The embodiment of the present application gradually expands neighboring nodes according to preset movement rules, which can reduce the search range of the next batch of nodes and significantly reduce the amount of calculation, effectively improving the efficiency of path search.

[0061] Optionally, in one embodiment of the present application, a preset heuristic function is used to perform cost calculation on each drivable node within a limited range to obtain a cost value of each drivable node, including: constructing a heuristic function based on the Euclidean distance cost from the current drivable node to the end point in the Cartesian coordinate system, the Euclidean distance cost between the target intelligent connected vehicle and the end point, and the road risk cost; based on a dynamic collision-free node set and a heuristic function, calculating the cost value of each drivable node within the limited range for each batch.

[0062] It is understandable that the heuristic function takes into account the straight-line distance from the node to the end point (distance factor), the road risk cost (risk factor), and the distance and angle from the node to the starting point, combined with the prior spatial relationship, but the latter does not directly participate in the cost calculation.

[0063] Specifically, the local path planning algorithm designed in the embodiment of the present application is similar to the heuristic search of A*, which can retrieve the position and risk of the dynamic collision-free node set and calculate the total cost of each path according to the heuristic function. Furthermore, for the sake of calculation speed, the embodiment of the present application can construct a Frenet coordinate system based on the central axis of the road for the selection of the path node of the vehicle, and design a heuristic function to guide the path search, wherein the embodiment of the present application can use the cost of the straight-line distance from the node to the end point on the Cartesian coordinate system and introduce the road risk cost as the heuristic function, and also use the distance and angle from the node to the starting point in combination with the prior spatial relationship as a prerequisite for finding adjacent nodes, without adding calculations, and the node with the lowest cost value will be set to the highest priority and added to the closed set. Optionally, in one embodiment of the present application, the expression of the heuristic function is:

[0064]

[0065] Among them, risk represents the cost value of each drivable node; dest(x, y) represents the coordinates of the destination; sample(x, y) and ICV(x, y) represent the coordinates of the sampling point and the target intelligent connected vehicle, respectively.

[0066] The embodiment of the present application constructs and utilizes a heuristic function to calculate the total cost of each potential driving path, comprehensively considering the safety and efficiency of driving, and can effectively avoid high-risk areas and significantly improve safety during driving.

[0067] Combination Figure 4 As shown, an embodiment of the present application is used to describe in detail the path planning method based on 4D risk occupancy under the vehicle-road-cloud collaborative framework.

[0068] It is understandable that the embodiments of the present application can integrate the resources and capabilities of ICV (vehicle), road-side sensing devices (road) and edge cloud computing platforms (cloud) to build an advanced intelligent transportation system architecture, that is, a vehicle-road-cloud collaborative framework, to provide comprehensive support for intelligent connected vehicles. Among them, the vehicle-road-cloud collaborative framework utilizes real-time data collected by intelligent connected vehicles and road-side sensing devices to achieve blind spot-free perception and beyond-line-of-sight perception, and combines the large-scale computing and storage capabilities of the cloud to achieve data integration and real-time processing, so that ICV can perform beyond-line-of-sight, blind spot-free, unified and intuitive dynamic path planning based on comprehensive and real-time environmental perception to prevent collision risks. In addition, the vehicle-road-cloud collaborative framework also has powerful data processing and analysis capabilities, can predict traffic trends, and respond to various changes in traffic conditions in a timely manner, which can significantly improve the operating efficiency and safety of the transportation system.

[0069] According to the path planning method based on 4D risk occupancy under the vehicle-road-cloud collaborative framework proposed in the embodiment of the present application, by obtaining the prior road information pre-stored by the edge cloud computing platform and the perception result data of multiple road-end perception units in the target area, the edge cloud computing platform obtains a potential path node set for the target intelligent connected vehicle based on the prior road information, the perception result data and the instruction information sent by the target intelligent connected vehicle; a risk value evaluation calculation is performed on each potential path node in the potential path node set to obtain a risk value evaluation result for each potential path node, and the initial drivable path nodes below the preset risk threshold are retained according to the risk value evaluation result to obtain a dynamic collision-free node set; based on the dynamic collision-free node set , the preset local path search strategy is used to limit the range of drivable nodes in each batch, and the preset heuristic function is used to calculate the cost of each drivable node within the limited range to obtain the cost value of each drivable node, so as to select each drivable node with the lowest cost value in the limited batch range as the optimal drivable node of the corresponding batch according to the cost value, and the final local path for the target intelligent connected vehicle is obtained based on the drivable nodes of each batch and the preset Gaussian filtering strategy, so as to effectively improve the computing efficiency by using the risk occupancy assessment algorithm and the local path planning algorithm and the vehicle-road-cloud collaborative architecture, alleviate the memory pressure brought by large-scale data, expand the perception range, and improve the adaptability in different traffic environments. As a result, the problems in related technologies such as high computational complexity, blind spot problems of occlusion of single-vehicle perception, and insufficient adaptability to specific scenarios, which limit the real-time, comprehensive and reliability of path planning, and may increase uncertainty and risk during driving, are alleviated.

[0070] Secondly, with reference to the accompanying drawings, a path planning device based on 4D risk occupancy under the vehicle-road-cloud collaborative framework proposed according to an embodiment of the present application is described.

[0071] Figure 5It is a block diagram of a path planning device based on 4D risk occupancy under the vehicle-road-cloud collaborative framework of an embodiment of the present application.

[0072] like Figure 5 As shown, the path planning device 10 based on 4D risk occupancy under the vehicle-road-cloud collaborative framework includes: an acquisition module 100, a risk assessment module 200 and a path planning module 300.

[0073] Specifically, the acquisition module 100 is used to obtain the prior road information pre-stored by the edge cloud computing platform and the perception result data of multiple road-end perception units in the target area, so that the edge cloud computing platform can obtain a potential path node set for the target intelligent connected vehicle based on the prior road information, perception result data and the instruction information sent by the target intelligent connected vehicle.

[0074] The risk assessment module 200 is used to perform risk value assessment calculation on each potential path node in the potential path node set, obtain the risk value assessment result of each potential path node, and retain the initial drivable path nodes below the preset risk threshold according to the risk value assessment result to obtain a dynamic collision-free node set.

[0075] The path planning module 300 is used to limit the range of the drivable nodes of each batch based on the dynamic collision-free node set by using a preset local path search strategy, and calculate the cost of each drivable node within the limited range by using a preset heuristic function to obtain the cost value of each drivable node, so as to select each drivable node with the lowest cost value within the limited batch range as the optimal drivable node of the corresponding batch according to the cost value, and obtain the final local path for the target intelligent connected vehicle based on the drivable nodes of each batch and the preset Gaussian filtering strategy.

[0076] Optionally, in one embodiment of the present application, the path planning module 300 includes: a determination unit, a retention unit, a first calculation unit and a filtering unit.

[0077] Among them, the determination unit is used to obtain the vehicle position of the target intelligent connected vehicle according to the command information sent by the target intelligent connected vehicle, so as to determine the initial neighboring nodes of the target intelligent connected vehicle through the vehicle position, and based on the initial neighboring nodes, determine the first batch of neighboring nodes in combination with the preset movement rules and the dynamic collision-free node set.

[0078] The retention unit is used to obtain the perception result data of each frame of the edge cloud computing platform and multiple road-side perception units, calculate the risk value of each drivable node based on the perception result data of each frame, and filter the risk value of each potential node to retain each drivable node below the preset risk threshold and construct a new dynamic collision-free node set.

[0079] The first calculation unit is used to limit the search range of each drivable node in each batch by using a local path search strategy, and calculate the cost value of each drivable node within the limited range by using a heuristic function, and select the drivable node with the smallest cost value as the optimal node of the corresponding batch.

[0080] The filtering unit is used to add the optimal nodes of each batch to the preset closed set in turn to obtain the optimal node closed set, and perform Gaussian filtering optimization operation on the optimal node closed set to obtain the final optimal local path node set, and based on the final optimal local path node set, generate the final local path for the target intelligent connected vehicle.

[0081] Optionally, in one embodiment of the present application, the path planning module 300 further includes: a construction unit and a second calculation unit.

[0082] Among them, the construction unit is used to construct a heuristic function according to the Euclidean distance cost from the current drivable node to the end point in the Cartesian coordinate system, the Euclidean distance cost between the target intelligent connected vehicle and the end point, and the road risk cost.

[0083] The second calculation unit is used to calculate the cost value of each drivable node in each batch within a limited range based on the dynamic collision-free node set and the heuristic function.

[0084] Optionally, in one embodiment of the present application, the mathematical expression of the heuristic function is:

[0085]

[0086] Among them, risk represents the cost value of each drivable node; dest(x, y) represents the coordinates of the destination; sample(x, y) and ICV(x, y) represent the coordinates of the sampling point and the target intelligent connected vehicle, respectively.

[0087] It should be noted that the above explanation of the embodiment of the path planning method based on 4D risk occupancy under the vehicle-road-cloud collaborative framework is also applicable to the path planning device based on 4D risk occupancy under the vehicle-road-cloud collaborative framework of this embodiment, and will not be repeated here.

[0088] According to the path planning device based on 4D risk occupancy under the vehicle-road-cloud collaborative framework proposed in the embodiment of the present application, it includes an acquisition module, which is used to obtain the prior road information pre-stored by the edge cloud computing platform and the perception result data of multiple road-end perception units in the target area, so that the edge cloud computing platform can obtain a potential path node set for the target intelligent networked vehicle according to the prior road information, the perception result data and the instruction information sent by the target intelligent networked vehicle; a risk assessment module, which is used to perform risk value assessment calculation on each potential path node in the potential path node set, obtain the risk value assessment result of each potential path node, and retain the initial drivable path nodes below the preset risk threshold according to the risk value assessment result to obtain a dynamic collision-free node set; a path planning module, It is used to limit the range of drivable nodes of each batch based on a dynamic collision-free node set using a preset local path search strategy, and calculate the cost of each drivable node within the limited range using a preset heuristic function to obtain the cost value of each drivable node, so as to select each drivable node with the lowest cost value within the limited batch range as the optimal drivable node of the corresponding batch according to the cost value, and obtain the final local path for the target intelligent connected vehicle based on the drivable nodes of each batch and the preset Gaussian filtering strategy, so as to effectively improve the computing efficiency by using the risk occupancy assessment algorithm and the local path planning algorithm as well as the vehicle-road-cloud collaborative architecture, alleviate the memory pressure brought by large-scale data, expand the perception range, and improve the adaptability in different traffic environments. As a result, the problems in related technologies such as high computational complexity, blind spot problems of occlusion of single-vehicle perception, and insufficient adaptability to specific scenarios, which limit the real-time, comprehensiveness and reliability of path planning, and may increase uncertainty and risk during driving, are alleviated.

[0089] Figure 6 A schematic diagram of the structure of a vehicle provided in an embodiment of the present application. The vehicle may include:

[0090] A memory 601 , a processor 602 , and a computer program stored in the memory 601 and executable on the processor 602 .

[0091] When the processor 602 executes the program, the path planning method based on 4D risk occupancy under the vehicle-road-cloud collaborative framework provided in the above embodiment is implemented.

[0092] Furthermore, the vehicle also includes:

[0093] The communication interface 603 is used for communication between the memory 601 and the processor 602 .

[0094] The memory 601 is used to store computer programs that can be executed on the processor 602 .

[0095] The memory 601 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0096] If the memory 601, the processor 602 and the communication interface 603 are implemented independently, the communication interface 603, the memory 601 and the processor 602 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0097] Optionally, in a specific implementation, if the memory 601, the processor 602 and the communication interface 603 are integrated on a chip, the memory 601, the processor 602 and the communication interface 603 can communicate with each other through an internal interface.

[0098] The processor 602 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0099] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned path planning method based on 4D risk occupancy under the vehicle-road-cloud collaborative framework.

[0100] An embodiment of the present application also provides a computer program product, which can run computer instructions, and when the computer instructions are executed by a processor, the above-mentioned path planning method based on 4D risk occupancy under the vehicle-road-cloud collaborative framework is implemented.

[0101] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0102] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0103] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or N executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0104] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or N wirings (electronic devices), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways as necessary and then storing it in a computer memory.

[0105] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above embodiment, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. If implemented by hardware, as in another embodiment, it can be implemented by any one or a combination of multiple of the following technologies known in the art: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0106] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0107] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0108] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A path planning method based on 4D risk occupancy in a vehicle-road-cloud collaborative framework, characterized in that: The following steps are involved: Acquire the prior road information pre-stored by the edge cloud computing platform and the perception result data of multiple road-end perception units in the target area, so that the edge cloud computing platform obtains a potential path node set for the target intelligent connected vehicle according to the prior road information, the perception result data and the instruction information sent by the target intelligent connected vehicle; Performing risk value evaluation calculation on each potential path node in the potential path node set to obtain a risk value evaluation result of each potential path node, and retaining initial drivable path nodes below a preset risk threshold according to the risk value evaluation result to obtain a dynamic collision-free node set; Based on the dynamic collision-free node set, a preset local path search strategy is used to limit the range of the drivable nodes in each batch, and a preset heuristic function is used to calculate the cost of each drivable node within the limited range to obtain a cost value of each drivable node, so as to select each drivable node with the lowest cost value within the limited batch according to the cost value as the optimal drivable node of the corresponding batch, and based on the drivable nodes of each batch and a preset Gaussian filtering strategy, a final local path for the target intelligent connected vehicle is obtained.

2. The method according to claim 1, characterized in that Based on the dynamic collision-free node set, the preset local path search strategy is used to limit the range of the drivable nodes of each batch, and the preset heuristic function is used to calculate the cost of each drivable node within the limited range to obtain the cost value of each drivable node, so as to select each drivable node with the lowest cost value within the limited batch range as the optimal drivable node of the corresponding batch according to the cost value, and the final local path facing the target intelligent connected vehicle is obtained based on the drivable nodes of each batch and the preset Gaussian filtering strategy, including: Acquire the vehicle position of the target intelligent connected vehicle according to the instruction information sent by the target intelligent connected vehicle, determine the initial neighboring nodes of the target intelligent connected vehicle through the vehicle position, and determine the first batch of neighboring nodes based on the initial neighboring nodes, combined with the preset movement rules and the dynamic collision-free node set; Acquire the perception result data of each frame of the edge cloud computing platform and the multiple road-side perception units, calculate the risk value of each drivable node according to the perception result data of each frame, and filter the risk value of each potential node to retain each drivable node below the preset risk threshold, and construct a new dynamic collision-free node set; The local path search strategy is used to limit the search range of each drivable node in each batch, and within the limited range, the cost value of each drivable node is calculated by the heuristic function, and the drivable node with the smallest cost value is selected as the optimal node of the corresponding batch; The optimal nodes of each batch are added to the preset closed set in turn to obtain the optimal node closed set, and the Gaussian filtering optimization operation is performed on the optimal node closed set to obtain the final optimal local path node set, and based on the final optimal local path node set, the final local path for the target intelligent connected vehicle is generated.

3. The method according to claim 1, characterized in that The cost calculation of each drivable node using a preset heuristic function within a limited range to obtain the cost value of each drivable node includes: Constructing the heuristic function according to the Euclidean distance cost from the current drivable node to the end point on the Cartesian coordinate system, the Euclidean distance cost between the target intelligent connected vehicle and the end point, and the road risk cost; Based on the dynamic collision-free node set and the heuristic function, a cost value of each drivable node in each batch within the limited range is calculated.

4. The method according to claim 3, characterized in that The expression of the heuristic function is: Among them, risk represents the cost value of each drivable node; dest(x, y) represents the coordinates of the destination; sample(x, y) and ICV(x, y) represent the coordinates of the sampling point and the target intelligent connected vehicle respectively.

5. A path planning device based on 4D risk occupancy in a vehicle-road-cloud collaborative framework, characterized in that: include: An acquisition module is used to acquire the prior road information pre-stored by the edge cloud computing platform and the perception result data of multiple road-end perception units in the target area, so that the edge cloud computing platform obtains a potential path node set for the target intelligent connected vehicle according to the prior road information, the perception result data and the instruction information sent by the target intelligent connected vehicle; A risk assessment module is used to perform risk value assessment calculation on each potential path node in the potential path node set to obtain a risk value assessment result of each potential path node, and retain the initial drivable path nodes below a preset risk threshold according to the risk value assessment result to obtain a dynamic collision-free node set; A path planning module is used to limit the range of drivable nodes in each batch based on the dynamic collision-free node set by using a preset local path search strategy, and calculate the cost of each drivable node within the limited range by using a preset heuristic function to obtain a cost value of each drivable node, so as to select each drivable node with the lowest cost value within the limited batch range as the optimal drivable node of the corresponding batch according to the cost value, and obtain the final local path for the target intelligent connected vehicle based on the drivable nodes of each batch and a preset Gaussian filtering strategy.

6. The device according to claim 5, characterized in that The path planning module includes: A determination unit, configured to obtain a vehicle position of the target intelligent connected vehicle according to the instruction information sent by the target intelligent connected vehicle, so as to determine an initial neighboring node of the target intelligent connected vehicle through the vehicle position, and determine a first batch of neighboring nodes based on the initial neighboring nodes, in combination with a preset movement rule and the dynamic collision-free node set; A retaining unit, used to obtain the perception result data of each frame of the edge cloud computing platform and the multiple road-side perception units, calculate the risk value of each drivable node according to the perception result data of each frame, and filter the risk value of each potential node to retain each drivable node below the preset risk threshold, and construct a new dynamic collision-free node set; A first calculation unit is used to limit the search range of each drivable node in each batch by using the local path search strategy, and calculate the cost value of each drivable node within the limited range by using the heuristic function, and select the drivable node with the smallest cost value as the optimal node of the corresponding batch; A filtering unit is used to add the optimal nodes of each batch to a preset closed set in turn to obtain an optimal node closed set, and perform a Gaussian filtering optimization operation on the optimal node closed set to obtain a final optimal local path node set, and based on the final optimal local path node set, generate the final local path for the target intelligent connected vehicle.

7. The device according to claim 5, characterized in that The path planning module also includes: A construction unit, configured to construct the heuristic function according to the Euclidean distance cost from the current drivable node to the end point on the Cartesian coordinate system, the Euclidean distance cost between the target intelligent connected vehicle and the end point, and the road risk cost; The second calculation unit is used to calculate the cost value of each drivable node in each batch within the limited range based on the dynamic collision-free node set and the heuristic function.

8. The device according to claim 7, characterized in that The mathematical expression of the heuristic function is: Among them, risk represents the cost value of each drivable node; dest(x, y) represents the coordinates of the destination; sample(x, y) and ICV(x, y) represent the coordinates of the sampling point and the target intelligent connected vehicle respectively.

9. A vehicle, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement a path planning method based on 4D risk occupancy under a vehicle-road-cloud collaborative framework as described in any one of claims 1 to 4.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement a path planning method based on 4D risk occupancy under a vehicle-road-cloud collaborative framework as described in any one of claims 1-4.

11. A computer program product, comprising a computer program, characterized in that The computer program is executed to implement the path planning method based on 4D risk occupancy under the vehicle-road-cloud collaborative framework as described in any one of claims 1-4.

Citation Information

Patent Citations

  • All-domain collaborative awareness and decision-making method and device based on vehicle and road cloud interface

    CN115116216A

  • Intelligent network connection automobile cooperation method and system based on risk situation cognition

    CN118430235A

  • Decision planning method and apparatus for unmanned vehicle, electronic device, and storage medium

    WO2024001393A1

Cited By

  • Generation method and device of 4D risk occupancy data set, equipment and medium

    CN120145638A

  • 4D risk occupancy dataset generation method and apparatus, device, and medium

    CN120145638B

  • Driving path planning method and device, electronic equipment, storage medium and program

    CN121660209A