Automatic driving path planning method, device and equipment and storage medium

Abstract

The invention provides an automatic driving path planning method and device, equipment and a storage medium, and relates to the technical field of automatic driving, and the method comprises the steps: obtaining an environment map and at least one piece of environment information, carrying out the meshing of the environment map, and obtaining a grid environment; extracting a virtual control potential field, an obstacle interference potential field and at least one environment potential field; fusing each potential field to generate a superposition control potential field; determining a starting point grid and an end point grid in the grid environment; searching the neighborhood grid with the minimum potential field value from the starting point grid until the ending point grid is searched, and obtaining a continuous grid path from the starting point grid to the ending point grid; wherein the potential field value of the neighborhood grid is calculated based on the superposition control potential field; and converting the continuous grid path into a longitude and latitude high coordinate point set and outputting the longitude and latitude high coordinate point set. According to the method, accurate path planning is realized through multi-source environment information fusion and grid modeling, and dependence on a high-precision map and a laser radar is reduced.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, specifically to an autonomous driving path planning method, apparatus, device, and storage medium. Background Technology

[0002] Artificial potential field algorithms are a common method in computer science, robotics, and automation. They simulate and plan robot motion by constructing an artificial potential field. In this field, the target location is typically assigned a lower potential energy, while obstacles are assigned a higher potential energy, thus guiding the robot to move along a path with decreasing potential energy.

[0003] Existing autonomous driving path planning algorithms for new energy vehicles largely rely on LiDAR and high-precision maps, making it difficult to operate efficiently in vision-driven, dynamic obstacle-prone, or unstructured road environments. This is especially true in the field of new energy light trucks, where the large inertia and turning radius of vehicles necessitate higher requirements for accuracy and safety in path planning. Traditional artificial potential field algorithms suffer from problems such as being prone to getting trapped in local optima and slow response speed, making it difficult to meet the dual requirements of real-time performance and stability.

[0004] Specifically, existing technologies face three core challenges: First, difficulty in understanding complex road environments. In unstructured or dynamically changing scenarios such as highways, mountainous areas, and urban-rural fringe areas, it is difficult to accurately model road topology and environmental features when lacking high-precision map support or relying solely on visual perception. Second, poor real-time performance and difficulty adapting to dynamic obstacles. Traditional global planning methods are slow to respond to sudden situations such as pedestrians crossing or vehicles cutting in, and cannot achieve rapid path replanning. Third, high computational overhead and difficulty in scaling. In large-scale, multi-obstacle scenarios, the planning process consumes a lot of computing power and has a slow response speed, making it difficult to adapt to resource-constrained edge computing environments such as onboard devices.

[0005] Therefore, there is an urgent need for an autonomous driving path planning solution that can operate efficiently in vision-dominated scenarios while taking into account real-time performance, safety, and low computing power requirements. Summary of the Invention

[0006] The purpose of this application is to provide an autonomous driving path planning method, device, electronic device, and storage medium to meet the requirements of high real-time performance and high safety for autonomous driving of vehicles.

[0007] In a first aspect, embodiments of this application provide an autonomous driving path planning method, including: Obtain an environmental map and at least one piece of environmental information, and use a preset grid modeling algorithm to grid the environmental map to obtain a grid environment, wherein each grid in the grid environment corresponds to a unique grid code; Extract virtual control potential field and obstacle interference potential field from the grid environment, and extract at least one environmental potential field from the at least one environmental information; The virtual control potential field, the obstacle interference potential field, and the at least one environmental potential field are fused together using a preset weighted superposition formula to generate a superimposed control potential field. Determine the start and end grids within the grid environment; Starting from the starting grid, a neighboring grid with the smallest potential field value is queried until the ending grid is found, resulting in a continuous grid path from the starting grid to the ending grid; wherein, the potential field value of the neighboring grid is calculated based on a cost calculation formula jointly determined by the potential energy change experienced from the starting grid to the current grid, the gravitational potential field of the ending grid on the current grid, and the superimposed control potential field; Based on the mapping relationship between grid encoding and latitude, longitude, and altitude coordinate systems, the continuous grid path is converted into a set of latitude, longitude, and altitude coordinate points and output.

[0008] In one possible implementation, the weighted superposition formula is as follows:

[0009] in, , , Indicates weight, Represents the superimposed control potential field. virtual For the virtual control potential field, Potential obs The potential field is the interference field of the obstacle. radar Represents an environmental potential field. It represents other environmental potential fields.

[0010] In one possible implementation, the method further includes: The weights are adjusted according to different driving tasks and / or environmental conditions.

[0011] In one possible implementation, the cost calculation formula is as follows:

[0012] in, This represents the potential field value of the current grid. This represents the potential energy change that has occurred from the starting grid to the current grid. This represents the gravitational potential field of the endpoint grid on the current grid. This represents the field strength control factor. This represents the heuristic intensity parameter.

[0013] In one possible implementation, converting the continuous grid path into a set of latitude, longitude, and elevation coordinate points and outputting it includes: Sparse operations are performed on the continuous grid path to preserve the set of grids at key inflection points; Based on the set of key inflection point grids, grids in the continuous grid path are added to the set of key inflection point grids at preset navigation point steps to obtain the target continuous grid path; The target continuous grid path is converted into a set of latitude, longitude, and altitude coordinate points and then output.

[0014] Secondly, embodiments of this application provide an autonomous driving path planning device, comprising: The acquisition module is used to acquire an environment map and at least one environmental information, and to grid the environment map using a preset grid modeling algorithm to obtain a grid environment, wherein each grid in the grid environment corresponds to a unique grid code; The overlay module is used to extract a virtual control potential field and an obstacle interference potential field from the grid environment, and to extract at least one environmental potential field from the at least one environmental information; and to fuse the virtual control potential field, the obstacle interference potential field and the at least one environmental potential field through a preset weighted overlay formula to generate an overlay control potential field. The determination module is used to determine the starting grid and the ending grid in the grid environment; The query module is used to query the neighboring grid with the smallest potential field value starting from the starting grid until the ending grid is found, thus obtaining a continuous grid path from the starting grid to the ending grid; wherein, the potential field value of the neighboring grid is calculated based on the cost calculation formula jointly determined by the potential field descent gradient from the starting grid to the current grid, the gravitational potential field of the ending grid on the current grid, and the superimposed control potential field. The conversion module is used to convert the continuous grid path into a set of latitude, longitude, and high-altitude coordinate points and output them according to the mapping relationship between the grid code and the latitude, longitude, and high-altitude coordinate system.

[0015] In one possible implementation, the weighted superposition formula is as follows:

[0016] in, , , Indicates weight, Represents the superimposed control potential field. virtual For the virtual control potential field, Potential obs The potential field is the interference field of the obstacle. radar Represents an environmental potential field. It represents other environmental potential fields.

[0017] In one possible implementation, the overlay module is further configured to: adjust the weights according to different driving tasks and / or environmental conditions.

[0018] In one possible implementation, the cost calculation formula is as follows:

[0019] in, This represents the potential field value of the current grid. This represents the potential field descent gradient from the starting grid to the current grid. This represents the gravitational potential field of the endpoint grid on the current grid. This represents the field strength control factor. This represents the heuristic intensity parameter.

[0020] In one possible implementation, the conversion module is specifically used for: Sparse operations are performed on the continuous grid path to preserve the set of grids at key inflection points; Based on the set of key inflection point grids, grids in the continuous grid path are added to the set of key inflection point grids at preset navigation point steps to obtain the target continuous grid path; The target continuous grid path is converted into a set of latitude, longitude, and altitude coordinate points and then output.

[0021] Thirdly, embodiments of this application provide an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the autonomous driving path planning method described in the first aspect of this application.

[0022] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-readable instructions thereon, which can be executed by a processor to implement the autonomous driving path planning method described in the first aspect of this application.

[0023] The autonomous driving path planning method, device, electronic device, and storage medium provided in this application acquire an environmental map and at least one piece of environmental information, and then use a preset grid modeling algorithm to grid the environmental map to obtain a grid environment. A virtual control potential field and an obstacle interference potential field are extracted from the grid environment, and at least one environmental potential field is extracted from the at least one piece of environmental information. The potential fields are fused using a preset weighted superposition formula to generate a superimposed control potential field. A starting grid and an ending grid are determined in the grid environment. Starting from the starting grid, a neighboring grid with the smallest potential field value is queried until the ending grid is found, resulting in a continuous grid path from the starting grid to the ending grid. The potential field value of the neighboring grid is calculated using a cost calculation formula determined by the potential energy change experienced from the starting grid to the current grid, the gravitational potential field of the ending grid on the current grid, and the superimposed control potential field. Based on the mapping relationship between the grid code and the latitude, longitude, and high-latitude coordinate system, the continuous grid path is converted into a set of latitude, longitude, and high-latitude coordinate points and output. Compared to existing technologies, this application quantifies complex environments into grid environments and combines multiple potential fields such as virtual control potential field, obstacle interference potential field, and environmental potential field to achieve accurate representation of complex, dynamic, and unstructured environments, thereby improving the autonomous driving system's understanding and adaptability to the environment. This application achieves accurate path planning through multi-source environmental information fusion and grid modeling, and also reduces the reliance on high-precision maps and LiDAR, thus promoting the application of low-cost autonomous driving solutions. Attached Figure Description

[0024] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart of an autonomous driving path planning method provided in this application is shown; Figure 2 A detailed flowchart of step S106 provided in this application is shown; Figure 3 A schematic diagram of the path planning algorithm provided in this application performing path exploration is shown; Figure 4 A schematic diagram of an autonomous driving path planning device provided in this application is shown. Detailed Implementation

[0025] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0026] It should be noted that, unless otherwise stated, the technical or scientific terms used in this application shall have the ordinary meaning as understood by one of ordinary skill in the art to which this application pertains.

[0027] Furthermore, the terms "first" and "second," etc., are used to distinguish different objects, not to describe a specific order. Additionally, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to those processes, methods, products, or apparatuses.

[0028] In order to achieve rapid control of the throttle valve pressure ratio or the pressure after the throttle valve and improve the controllability of engine torque, embodiments of this application provide an autonomous driving path planning method and device, an electronic device, and a computer-readable storage medium, which will be described below with reference to the accompanying drawings.

[0029] Please refer to Figure 1 The diagram illustrates a flowchart of an autonomous driving path planning method provided in this application. Figure 1 As shown, the method includes the following steps: S101. Obtain an environmental map and at least one piece of environmental information, and use a preset grid modeling algorithm to grid the environmental map to obtain a grid environment, wherein each grid in the grid environment corresponds to a unique grid code; The executing entity in this application embodiment can be the vehicle's autonomous driving system. The environmental map can be an urban area environmental map obtained by the vehicle through a network, including information such as buildings and roads. The environmental information can be environmental information collected by sensors such as onboard millimeter-wave radar and cameras, including environmental influencing factors such as the position and speed of surrounding vehicles, road boundaries, information on moving obstacles, road sign information, weather information, and road surface smoothness.

[0030] The preset mesh modeling algorithm is preferably the GeoSOT-3D subdivision algorithm, but DGGS (Discrete Global Grid Systems), quadtree mesh algorithm, or other equivalent mesh modeling algorithms can also be used. The mesh modeling algorithm quantifies the three-dimensional spatial environment into mesh cells. Each mesh cell carries environmental attribute information corresponding to its spatial location, and a unique mesh code enables precise positioning and management of the mesh cells, providing a foundation for subsequent potential field extraction and path search.

[0031] For example, the GeoSOT-3D subdivision algorithm can be used to grid the environmental map, resulting in a grid environment. Each grid cell in this environment contains spatial coordinates (corresponding to latitude, longitude, and altitude) and a unique grid code. The grid size can be set to 10 meters × 10 meters × 5 meters.

[0032] S102. Extract the virtual control potential field and obstacle interference potential field from the grid environment, and extract at least one environmental potential field from the at least one environmental information. The core of the virtual control potential field is to transform artificially set control rules, such as vehicle performance constraints (e.g., turning radius, inertia constraints) and driving rules (e.g., speed limits, lane priorities), into quantifiable potential field values ​​that guide the path to meet safe driving characteristics. The core of the obstacle interference potential field is to transform obstacles (static / dynamic) in the grid environment into repulsive potential fields to prevent collisions between the path and obstacles. During the extraction process, calculations must be strictly based on the grid coordinates and codes to ensure a one-to-one correspondence between each potential field value and the grid code, facilitating the subsequent fusion of the control potential fields.

[0033] The core of extracting environmental potential fields from environmental information is to transform unstructured environmental perception data (such as radar, vision, terrain, and meteorology) into quantifiable and computable numerical potential fields. Each potential field corresponds to a specific environmental influence dimension. For example, the three most commonly used environmental potential fields in autonomous driving scenarios are: environmental radar envelope field, terrain adaptation potential field, and dynamic risk potential field. The extraction of all environmental potential fields is based on the constructed grid environment, and each potential field value corresponds one-to-one with the grid code.

[0034] The core of the environmental radar envelope field is to transform spatial risks (such as obstacle distance, degree of obstruction, and dynamic target movement trend) within the radar coverage area into a potential field in the form of a heat map based on the detection data of vehicle-mounted radar (millimeter-wave radar, lidar).

[0035] The core of terrain-adaptive potential field is to transform terrain information (slope, undulation, ground hardness) into a potential field to guide the path to avoid terrain-restricted areas (such as steep slopes and potholes).

[0036] The core of the dynamic risk potential field is to convert the grid that may be occupied by dynamic obstacles such as pedestrians, non-motorized vehicles, and other vehicles in the future into a potential field based on the predicted movement trajectory, thereby avoiding collision risks in advance. Unlike the aforementioned static obstacle interference potential field, this dynamic risk potential field focuses on predictive risks.

[0037] After all potential fields are extracted, they need to be normalized to the [0,1] interval to avoid the difference in dimensions affecting the subsequent weighted fusion effect. 0 indicates no effect, and 1 indicates the greatest impact.

[0038] S103. The virtual control potential field, the obstacle interference potential field, and the at least one environmental potential field are fused using a preset weighted superposition formula to generate a superimposed control potential field. The weighted summation formula is as follows:

[0039] in, , , Indicates weight, Represents the superimposed control potential field. virtual For the virtual control potential field, Potential obs The potential field is the interference field of the obstacle. radar This represents an environmental potential field (specifically, the environmental radar envelope field). It represents other environmental potential fields (such as terrain-adaptive potential fields, dynamic risk potential fields, etc.).

[0040] Weight , , The weights of the virtual control potential field and other environmental potential fields can be dynamically adjusted according to different driving tasks (such as high-speed driving and low-speed obstacle avoidance) and / or environmental conditions (such as sunny days, rainy days, congested road sections, and open road sections) to achieve path strategy preference control in different scenarios. For example, in high-speed driving scenarios, the weight of the virtual control potential field can be increased to ensure a smooth path that conforms to the characteristics of high-speed vehicle driving; in congested road sections, the weights of the obstacle interference potential field and the environmental radar envelope field can be increased to improve the obstacle avoidance flexibility of the path.

[0041] S104. Determine the starting grid and the ending grid in the grid environment; The starting grid is the grid corresponding to the current location of the autonomous driving device (such as a new energy light truck), and the ending grid is the grid corresponding to the target location of the autonomous driving device. The latitude, longitude, and altitude coordinates of the starting and ending points can be obtained through a positioning system (such as GPS or BeiDou positioning), and then the corresponding starting and ending grids are determined according to the mapping relationship between the grid code and the latitude, longitude, and altitude coordinate system. For example, the code for the starting grid corresponding to the vehicle's current location is determined to be G001, and the code for the ending grid corresponding to the destination is determined to be G099.

[0042] S105. Starting from the starting grid, query the neighboring grid with the smallest potential field value until the ending grid is found, thus obtaining a continuous grid path from the starting grid to the ending grid; wherein, the potential field value of the neighboring grid is calculated based on the cost calculation formula jointly determined by the potential energy change experienced from the starting grid to the current grid, the gravitational potential field of the ending grid on the current grid, and the superimposed control potential field. The formula for calculating the above cost is as follows:

[0043] in, This represents the potential field value of the current grid. This represents the potential energy change that has occurred from the starting grid to the current grid. This represents the gravitational potential field of the endpoint grid on the current grid. This represents the field strength control factor. This represents the heuristic strength parameter. In the process of searching for a reasonable path, this application uses a heuristic search algorithm with a greedy strategy, and the heuristic strength parameter is a parameter in this algorithm.

[0044] Specifically, The potential energy change that has been experienced is represented by grid distance, with the value being larger the closer to the endpoint, which is used to measure the degree of potential energy decrease that has been experienced. The cost of the superimposed control potential field in the grid environment is represented by a normalized form. The gravitational potential field of the endpoint relative to the current node decreases in value as it approaches the endpoint, with the gradient pointing towards the endpoint. Therefore, this application extracts all elements that cause interference in vehicle path planning from the grid environment and performs a unified weighted expression, making all elements affecting the path a component of the superimposed control potential field, forming a comprehensive control potential field that adapts to vehicle path performance characteristics and external environmental factors. The vehicle path trajectory moves along the gradient direction of the comprehensive control potential field, ultimately pointing towards the endpoint.

[0045] Specifically, a 26-neighborhood query method can be used to set the field strength control factor. =0.8, heuristic intensity parameter =0.5, and the cost calculation formula is used to iteratively search for the neighborhood grid with the smallest potential field value starting from the starting grid G001 until the ending grid G099 is reached, thus obtaining a continuous grid path.

[0046] The preferred neighborhood grid query method is the 26-neighbor query method, which queries the 26 neighboring grids of the current grid in the three-dimensional space in the vertical, horizontal, front-back and diagonal directions. This ensures the omnidirectionality and continuity of the path search and avoids the path getting stuck in a local optimum due to the limitation of the neighborhood range.

[0047] After generating the aforementioned superimposed control potential field, this application uses it as the path cost and employs a heuristic guidance strategy. Based on the aforementioned superimposed control potential field, a gravitational potential field is then combined... and potential field change value The optimal path from the starting point to the destination is planned by searching for the direction of gradient descent in the potential field. This search process takes into account all environmental and navigational constraints, ensuring that the found path is both safe and efficient.

[0048] The heuristic guidance strategy described in this application can be replaced by other optimization algorithms depending on the specific nature and requirements of the problem. These alternative algorithms may include problem-specific heuristic strategies, metaheuristic algorithms, exact algorithms, or hybrid strategies.

[0049] S106. Based on the mapping relationship between the grid code and the latitude, longitude and altitude coordinate system, the continuous grid path is converted into a set of latitude, longitude and altitude coordinate points and output.

[0050] Specifically, such as Figure 2 As shown, step S106 above can be achieved as follows: S201. Perform a sparse operation on the continuous grid path to preserve the set of key inflection point grids; A sparse method could be the Douglas-Puk algorithm, which works by recursively calculating the perpendicular distance from each point on the curve to the line connecting the starting and ending points. Points with distances greater than a threshold (those contributing most to the path shape) are retained, while points with distances less than the threshold (redundant points) are removed, thus simplifying the curve. As can be seen, the key inflection point grid consists of grids where the path direction changes significantly; sparse operations can remove redundant grid nodes from the path, simplifying the path data.

[0051] S202. Based on the key inflection point grid set, the grids in the continuous grid path are added to the key inflection point grid set at preset navigation point steps to obtain the target continuous grid path. The preset navigation point step size can be set according to the actual application scenario, preferably D kilometers (D is a real number greater than 0, such as 0.2 kilometers or 0.5 kilometers), or it can be set to meter-level step size according to the path accuracy requirements and the response capability of the autonomous driving equipment. The smoothness and accuracy of the path are ensured by supplementing grid nodes.

[0052] S203. Convert the target continuous grid path into a set of latitude, longitude, and high-order coordinate points and output it. According to the preset mapping relationship between the grid code and the latitude, longitude, and high-order coordinate system, extract the latitude, longitude, and high-order coordinates of the center point of each grid to form a set of latitude, longitude, and high-order coordinate points. This set of coordinate points can be directly used as the navigation path input for the autonomous driving device.

[0053] For ease of understanding, this application also provides the following specific path planning algorithm: Step 1: Determine the starting speed of the neighborhood query ; From the starting grid Begin a 26-neighborhood search. Based on the previously experienced potential energy changes... Superimposed control potential field and gravitational potential field Under the combined effect, the grid with the smallest overall estimated potential field value is selected. As an action strategy, the grid is defined as And record the grid code:

[0054] Step2: Judgment With the end grid Are they equal? ​​If not, As a new Return to Step 1; if they are equal, proceed to Step 3. Step 3: Output the data recorded at this point. A set of grid codes, output as continuous grid paths under the current parameter state:

[0055] Step 4: [The sentence is incomplete and requires more context.] Perform sparsity optimization while preserving the grid-coded set of key inflection points. ; Step 5: Based on the key inflection points, every D kilometers (navigation point step size) Add the points in the middle In this process, the path grid encoding set is obtained. ; Step 6: According to Based on the mapping relationship between grid coding and latitude / longitude coordinate systems, the center point is obtained. , which serves as the set of path input coordinates received by the vehicle;

[0056] Step 7: Output the path grid code set and vehicle navigation point latitude, longitude and altitude coordinate set The algorithm is complete.

[0057] like Figure 3 The diagram shown illustrates path exploration based on the aforementioned path planning algorithm. Figure 3 The search begins at the starting grid, proceeds through the discovered path grids, and reaches the ending grid. Obstacle grids in the diagram represent the presence of obstacles, while free grids represent the absence of obstacles.

[0058] The autonomous driving path planning method provided in this application quantifies complex environments into grid environments through a grid modeling algorithm. By combining multiple potential fields, including virtual control potential field, obstacle interference potential field, and environmental potential field, it achieves accurate representation of complex, dynamic, and unstructured environments, improving the autonomous driving system's understanding and adaptability to the environment. The path search is guided by neighborhood query and cost calculation formulas, effectively avoiding the problem of traditional algorithms easily getting trapped in local optima, ensuring the optimality and continuity of the path. Through a dynamic weight adjustment mechanism, it can adapt to different driving tasks and environmental conditions, improving the flexibility and practicality of path planning. Grid encoding enables efficient management and calculation of environmental information, reducing the computational overhead of path planning and meeting the needs of resource-constrained edge computing environments such as onboard equipment. Accurate path planning through multi-source environmental information fusion and grid modeling also reduces reliance on high-precision maps and LiDAR, promoting the application of low-cost autonomous driving solutions.

[0059] In the above embodiments, an autonomous driving path planning method is provided. Correspondingly, this application also provides an autonomous driving path planning device, which can be implemented by software, hardware, or a combination of both. For example, the autonomous driving path planning device may include integrated or separate functional modules or units to perform the corresponding steps in the above methods. Please refer to... Figure 4 This illustration shows a schematic diagram of an autonomous driving path planning device provided by some embodiments of this application. Since the device embodiments are basically similar to the method embodiments, the description is relatively simple; relevant details can be found in the description of the method embodiments. The device embodiments described below are merely illustrative.

[0060] like Figure 4 As shown, the autonomous driving path planning device 10 may include: The acquisition module 101 is used to acquire an environmental map and at least one environmental information, and to grid the environmental map using a preset grid modeling algorithm to obtain a grid environment, wherein each grid in the grid environment corresponds to a unique grid code; The overlay module 102 is used to extract a virtual control potential field and an obstacle interference potential field from the grid environment, and to extract at least one environmental potential field from the at least one environmental information; and to fuse the virtual control potential field, the obstacle interference potential field and the at least one environmental potential field through a preset weighted overlay formula to generate an overlay control potential field. The determination module 103 is used to determine the starting grid and the ending grid in the grid environment; The query module 104 is used to query the neighboring grid with the smallest potential field value starting from the starting grid until the ending grid is found, so as to obtain a continuous grid path from the starting grid to the ending grid; wherein, the potential field value of the neighboring grid is calculated based on the cost calculation formula jointly determined by the potential field descent gradient from the starting grid to the current grid, the gravitational potential field of the ending grid on the current grid, and the superimposed control potential field. The conversion module 105 is used to convert the continuous grid path into a set of latitude, longitude and high coordinate points and output it according to the mapping relationship between the grid code and the latitude, longitude and high coordinate system.

[0061] In one possible implementation, the weighted superposition formula is as follows:

[0062] in, , , Indicates weight, Represents the superimposed control potential field. virtual For the virtual control potential field, Potential obs The potential field is the interference field of the obstacle. radar Represents an environmental potential field. It represents other environmental potential fields.

[0063] In one possible implementation, the superposition module 102 is further configured to: adjust the weights according to different driving tasks and / or environmental conditions.

[0064] In one possible implementation, the cost calculation formula is as follows:

[0065] in, This represents the potential field value of the current grid. This represents the potential field descent gradient from the starting grid to the current grid. This represents the gravitational potential field of the endpoint grid on the current grid. This represents the field strength control factor. This represents the heuristic intensity parameter.

[0066] In one possible implementation, the conversion module 105 is specifically used for: Sparse operations are performed on the continuous grid path to preserve the set of grids at key inflection points; Based on the set of key inflection point grids, grids in the continuous grid path are added to the set of key inflection point grids at preset navigation point steps to obtain the target continuous grid path; The target continuous grid path is converted into a set of latitude, longitude, and altitude coordinate points and then output.

[0067] The autonomous driving path planning device provided in this application quantifies complex environments into grid environments and combines multiple potential fields such as virtual control potential field, obstacle interference potential field, and environmental potential field to achieve accurate representation of complex, dynamic, and unstructured environments, thereby improving the autonomous driving system's understanding and adaptability to the environment. This application achieves accurate path planning through multi-source environmental information fusion and grid modeling, and also reduces the reliance on high-precision maps and LiDAR, thereby promoting the application of low-cost autonomous driving solutions.

[0068] This application also provides an electronic device corresponding to the autonomous driving path planning method provided in the foregoing embodiments. The electronic device may be an in-vehicle device, a mobile phone, a laptop, a tablet computer, a desktop computer, etc., to execute the above-mentioned autonomous driving path planning method.

[0069] The electronic device provided in this application embodiment and the autonomous driving path planning method provided in this application embodiment are based on the same inventive concept and have the same beneficial effects as the methods they adopt, operate or implement.

[0070] This application also provides a computer-readable storage medium corresponding to the autonomous driving path planning method provided in the foregoing embodiments, wherein a computer program (i.e., a program product) is stored thereon, and the computer program, when run by a processor, executes the autonomous driving path planning method provided in any of the foregoing embodiments.

[0071] It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical and magnetic storage media, which will not be elaborated here.

[0072] The computer-readable storage medium provided in the above embodiments of this application and the autonomous driving path planning method provided in the embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the applications stored therein.

[0073] It should be noted that the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0074] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and they should all be covered within the scope of the claims and specification of this application.

Claims

1. An autonomous driving path planning method, characterized in that, include: Obtain an environmental map and at least one piece of environmental information, and use a preset grid modeling algorithm to grid the environmental map to obtain a grid environment, wherein each grid in the grid environment corresponds to a unique grid code; Extract virtual control potential field and obstacle interference potential field from the grid environment, and extract at least one environmental potential field from the at least one environmental information; The virtual control potential field, the obstacle interference potential field, and the at least one environmental potential field are fused together using a preset weighted superposition formula to generate a superimposed control potential field. Determine the start and end grids within the grid environment; Starting from the starting grid, a neighboring grid with the smallest potential field value is queried until the ending grid is found, resulting in a continuous grid path from the starting grid to the ending grid; wherein, the potential field value of the neighboring grid is calculated based on a cost calculation formula jointly determined by the potential energy change experienced from the starting grid to the current grid, the gravitational potential field of the ending grid on the current grid, and the superimposed control potential field; Based on the mapping relationship between grid encoding and latitude, longitude, and altitude coordinate systems, the continuous grid path is converted into a set of latitude, longitude, and altitude coordinate points and output.

2. The autonomous driving path planning method according to claim 1, characterized in that, The weighted superposition formula is as follows: in, , , Indicates weight, Represents the superimposed control potential field. virtual For the virtual control potential field, Potential obs The potential field is the interference field of the obstacle. radar Represents an environmental potential field. It represents other environmental potential fields.

3. The autonomous driving path planning method according to claim 2, characterized in that, The method further includes: The weights are adjusted according to different driving tasks and / or environmental conditions.

4. The autonomous driving path planning method according to claim 2, characterized in that, The cost calculation formula is as follows: in, This represents the potential field value of the current grid. This represents the potential energy change that has occurred from the starting grid to the current grid. This represents the gravitational potential field of the endpoint grid on the current grid. This represents the field strength control factor. This represents the heuristic intensity parameter.

5. The autonomous driving path planning method according to claim 1, characterized in that, The step of converting the continuous grid path into a set of latitude, longitude, and elevation coordinate points and outputting it includes: Sparse operations are performed on the continuous grid path to preserve the set of grids at key inflection points; Based on the set of key inflection point grids, grids in the continuous grid path are added to the set of key inflection point grids at preset navigation point steps to obtain the target continuous grid path; The target continuous grid path is converted into a set of latitude, longitude, and altitude coordinate points and then output.

6. An autonomous driving path planning device, characterized in that, include: The acquisition module is used to acquire an environment map and at least one environmental information, and to grid the environment map using a preset grid modeling algorithm to obtain a grid environment, wherein each grid in the grid environment corresponds to a unique grid code; The overlay module is used to extract a virtual control potential field and an obstacle interference potential field from the grid environment, and to extract at least one environmental potential field from the at least one environmental information; and to fuse the virtual control potential field, the obstacle interference potential field and the at least one environmental potential field through a preset weighted overlay formula to generate an overlay control potential field. The determination module is used to determine the starting grid and the ending grid in the grid environment; The query module is used to query the neighboring grid with the smallest potential field value starting from the starting grid until the ending grid is found, thus obtaining a continuous grid path from the starting grid to the ending grid; wherein, the potential field value of the neighboring grid is calculated based on the cost calculation formula jointly determined by the potential field descent gradient from the starting grid to the current grid, the gravitational potential field of the ending grid on the current grid, and the superimposed control potential field. The conversion module is used to convert the continuous grid path into a set of latitude, longitude, and high-altitude coordinate points and output them according to the mapping relationship between the grid code and the latitude, longitude, and high-altitude coordinate system.

7. The autonomous driving path planning device according to claim 6, characterized in that, The weighted superposition formula is as follows: in, , , Indicates weight, Represents the superimposed control potential field. virtual For the virtual control potential field, Potential obs The potential field is the interference field of the obstacle. radar Represents an environmental potential field. It represents other environmental potential fields.

8. The autonomous driving path planning device according to claim 7, characterized in that, The cost calculation formula is as follows: in, This represents the potential field value of the current grid. This represents the potential field descent gradient from the starting grid to the current grid. This represents the gravitational potential field of the endpoint grid on the current grid. This represents the field strength control factor. This represents the heuristic strength parameter.

9. An electronic device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor, when executing the computer program, implements the autonomous driving path planning method as described in any one of claims 1 to 5.

10. A computer-readable storage medium, characterized in that, It stores computer-readable instructions that can be executed by a processor to implement the autonomous driving path planning method as described in any one of claims 1 to 5.

Images