Path planning method and device based on obstacle modification, equipment and storage medium

Abstract

The invention relates to the technical field of automatic driving path planning, and discloses a path planning method, device and equipment based on obstacle modification, and a storage medium, which are used for improving the safety and operation efficiency of automatic driving path planning. The path planning method based on obstacle modification comprises the following steps: acquiring a modification instruction of a user on a static obstacle of a mine through a visual interface, further generating standardized update information with priority, constructing a space-time cost map in combination with mine environment data, vehicle kinematics parameters and a dynamic obstacle prediction trajectory, and performing path planning according to the space-time cost map. And performing path search based on the map to generate a global reference path, and cooperatively issuing the global reference path to the vehicle end through the vehicle cloud to guide the vehicle to run safely and efficiently.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving path planning technology, and in particular to a path planning method, apparatus, device, and storage medium based on obstacle modification. Background Technology

[0002] In today's era of rapid development of automation technology, mining operations are accelerating towards intelligence and automation. Autonomous vehicles, with their advantages of high efficiency, stability, and ability to operate continuously for long periods of time, are being used more and more widely in mining environments.

[0003] Currently, mining operations typically involve environments characterized by significant unstructured features, meaning a lack of fixed road planning, random and variable shapes and locations of obstacles (such as rocks, material piles, and equipment), and complex and dynamically evolving terrain. In this environment, existing autonomous driving path planning methods largely rely on pre-set fixed rules and static map data, lacking effective mechanisms for handling the aforementioned dynamic environment. Current technologies usually rely on vehicle-side perception for local real-time obstacle avoidance, but this method has significant limitations: vehicle-side sensors have limited perception range, resulting in numerous blind spots in complex terrain; furthermore, insufficient computing resources make it difficult to perform sufficiently safe and smooth global path replanning for static obstacles suddenly appearing in distant or hidden areas, causing vehicles to fail to avoid obstacles in time. This not only severely impacts operational efficiency but also poses a safety hazard of collisions. Summary of the Invention

[0004] This invention provides a path planning method, apparatus, device, and storage medium based on obstacle modification, to solve the problem that the existing technology relies solely on vehicle-side perception for path planning in the mining environment, which leads to reduced work efficiency and significant safety hazards.

[0005] The first aspect of this invention provides a path planning method based on obstacle modification, comprising: obtaining an operation instruction from a user on a visual interactive interface to modify static obstacle information in an unstructured mining scene; generating standardized static obstacle update information according to the operation instruction, and prioritizing each static obstacle update information to obtain target static obstacle update information, wherein one static target obstacle update information corresponds to one static obstacle; acquiring mining environment data, the vehicle's own equipment kinematic parameters, and the future predicted trajectories of multiple dynamic obstacles; constructing a spatiotemporal cost map based on the target static obstacle update information, the mining environment data, the equipment kinematic parameters, and the future predicted trajectories of the multiple dynamic obstacles; searching for a global reference path based on the spatiotemporal cost map using a path planning algorithm; and distributing the global reference path to the vehicle terminal so that the vehicle terminal controls vehicle operation based on the global reference path.

[0006] In one feasible implementation, obtaining the user's operation instructions includes: receiving an operation instruction from the user to add a new static obstacle using a graphical drawing tool on the visual interactive interface, the instruction including the geometric shape, size, and position information of the new static obstacle; receiving an instruction from the user to modify the attributes of an existing static obstacle through dragging, scaling, or rotating operations, the attributes including the position, size, orientation, and type of the static obstacle; receiving an instruction from the user to remove a specified static obstacle by selecting and confirming a delete operation; and encapsulating the add, modify, and delete operations into a structured event data stream.

[0007] In one feasible implementation, the step of generating standardized obstacle update information according to the operation instructions and prioritizing each obstacle update information to obtain target obstacle update information includes: parsing the operation instructions, identifying the operation type and corresponding static obstacle parameters, wherein the operation type includes adding, modifying, or deleting static obstacles; encapsulating the parsed parameters into standardized static obstacle update information, wherein the standardized format includes a unique identifier for the static obstacle, a type attribute, a set of geometric parameters, and a timestamp; and assigning a corresponding priority tag to each static obstacle update information based on the type attribute of the static obstacle and the importance of its location area to form target static obstacle update information.

[0008] In one feasible implementation, constructing a spatiotemporal cost map based on the updated information of each target static obstacle, the mine environment data, the equipment kinematic parameters, and the future predicted trajectories of the multiple dynamic obstacles includes: fusing the updated information of each target static obstacle into the mine environment data according to its priority using different cost decay functions to generate enhanced static environment data; calculating the braking distance and turning radius of the vehicle based on the equipment kinematic parameters and converting them into safe space constraints; and generating a spatiotemporal cost map based on the enhanced static environment data and the safe space constraints, combined with the future predicted trajectories of the multiple dynamic obstacles.

[0009] In one feasible implementation, the step of fusing the update information of each target obstacle into the mine environment data according to its priority and using different cost decay functions to generate enhanced static environment data includes: assigning a corresponding cost decay coefficient to the update information of each target static obstacle according to its priority, with high-priority static obstacles using a slow decay function and low-priority static obstacles using a fast decay function; performing nonlinear superposition calculation on the cost value of the grid where the static obstacle is located and its affected area according to the assigned decay function; and fusing the superimposed cost value with the mine environment data to generate enhanced static environment data.

[0010] In one feasible implementation, calculating the braking distance and turning radius of the vehicle based on the device's kinematic parameters and converting them into safety space constraints includes: calculating the minimum safe braking distance at different speeds based on the vehicle's maximum deceleration, current speed, and system reaction delay time in the device's kinematic parameters; calculating the minimum turning radius of the vehicle at different speeds based on the vehicle's wheelbase and maximum steering angle in the device's kinematic parameters; mapping the minimum safe braking distance and minimum turning radius to an environmental space model to generate a safety buffer area outside the vehicle's outline, and using the safety buffer area as a safety space constraint during path search.

[0011] In one feasible implementation, generating a spatiotemporal cost map based on the enhanced static environmental data and the safety space constraints, combined with the future predicted trajectories of the multiple dynamic obstacles, includes: constructing an initial two-dimensional environmental cost map based on the enhanced static environmental data; preprocessing the two-dimensional environmental cost map using the safety space constraints as a passability condition to mark areas that do not conform to vehicle kinematic constraints; discretizing the future predicted trajectories of the multiple dynamic obstacles in the time dimension to generate a series of dynamic obstacle occupancy grids with time steps to obtain dynamic obstacle information; and fusing the dynamic obstacle information with the preprocessed two-dimensional environmental cost map in the spatiotemporal dimension to generate a three-dimensional spatiotemporal cost map.

[0012] A second aspect of the present invention provides a path planning device based on obstacle modification, comprising: a first acquisition module, configured to acquire operation instructions from a user on a visual interactive interface to modify static obstacle information in an unstructured mining scene; a processing module, configured to generate standardized static obstacle update information according to the operation instructions, and to prioritize each static obstacle update information to obtain target static obstacle update information, wherein one target static obstacle update information corresponds to one obstacle; a second acquisition module, configured to acquire mining environment data and the vehicle's own equipment kinematic parameters; a construction module, configured to fuse the target static obstacle update information, the mining environment data, and the equipment kinematic parameters, and to predict the future trajectories of multiple dynamic obstacles to construct a spatiotemporal cost map; a search module, configured to search based on the spatiotemporal cost map using a path planning algorithm to obtain a global reference path; and a distribution module, configured to distribute the global reference path to the vehicle terminal, so that the vehicle terminal controls the vehicle operation based on the global reference path.

[0013] In one feasible implementation, the first acquisition module is specifically configured to: receive an operation instruction from a user on the visual interactive interface to add a new static obstacle using a graphical drawing tool, the instruction including the geometric shape, size, and position information of the new static obstacle; receive an instruction from a user to modify the attributes of an existing static obstacle through dragging, scaling, or rotating operations, the attributes including the position, size, orientation, and type of the static obstacle; receive an instruction from a user to remove a specified static obstacle by selecting and confirming a delete operation; and encapsulate the add, modify, and delete operations into a structured event data stream.

[0014] In one feasible implementation, the processing module is specifically used to: parse the operation instruction, identify the operation type and the corresponding static obstacle parameters, wherein the operation type includes adding, modifying or deleting static obstacles; encapsulate the parsed parameters into static obstacle update information in a standardized format, wherein the standardized format includes a unique identifier for the static obstacle, a type attribute, a set of geometric parameters and a timestamp; and assign a corresponding priority marker to each static obstacle update information based on the type attribute of the static obstacle and the importance of its location area, thereby forming target static obstacle update information.

[0015] In one feasible implementation, the construction module includes: a first generation unit, configured to integrate the update information of each target static obstacle into the mine environment data according to its priority using different cost decay functions to generate enhanced static environment data; a processing unit, configured to calculate the braking distance and turning radius of the vehicle based on the equipment kinematic parameters and convert them into safe space constraints; and a second generation unit, configured to generate a spatiotemporal cost map based on the enhanced static environment data and the safe space constraints, combined with the future predicted trajectories of the multiple dynamic obstacles. In one feasible implementation, the first generation unit is specifically used to: assign a corresponding cost attenuation coefficient to the update information of each target static obstacle according to its priority, with high-priority static obstacles using a slow attenuation function and low-priority static obstacles using a fast attenuation function; perform nonlinear superposition calculation on the cost value of the grid where the static obstacle is located and its affected area according to the assigned attenuation function; and fuse the superimposed cost value with the mine environment data to generate enhanced static environment data. In one feasible implementation, the processing unit is specifically used to: calculate the minimum safe braking distance at different speeds based on the vehicle's maximum deceleration, current speed, and system reaction delay time in the device's kinematic parameters; calculate the minimum turning radius of the vehicle at different speeds based on the vehicle's wheelbase and maximum steering angle in the device's kinematic parameters; map the minimum safe braking distance and minimum turning radius into an environmental space model to generate a safe buffer area outside the vehicle's outline, and use the safe buffer area as a safe space constraint during path search. In one feasible implementation, the second generation unit is specifically used to: construct an initial two-dimensional environmental cost map based on the enhanced static environmental data; preprocess the two-dimensional environmental cost map using the safe space constraint as a passability condition, and mark areas that do not conform to vehicle kinematic constraints; discretize the future predicted trajectories of the multiple dynamic obstacles in the time dimension to generate a series of dynamic obstacle occupancy grids with time steps, thereby obtaining dynamic obstacle information; and fuse the dynamic obstacle information with the preprocessed two-dimensional environmental cost map in the spatiotemporal dimension to generate a three-dimensional spatiotemporal cost map.

[0016] A third aspect of the present invention provides an electronic device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor invokes the instructions in the memory to cause the electronic device to perform the obstacle-modification-based path planning method described above.

[0017] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the above-described obstacle-modification-based path planning method.

[0018] The technical solution provided by this invention involves: obtaining user operation instructions on a visual interactive interface to modify static obstacle information in an unstructured mining scene; generating standardized static obstacle update information based on the operation instructions, and prioritizing each static obstacle update information to obtain target static obstacle update information, with one static target obstacle update information corresponding to one static obstacle; acquiring mining environment data, vehicle's own equipment kinematic parameters, and the future predicted trajectories of multiple dynamic obstacles; constructing a spatiotemporal cost map based on the target static obstacle update information, the mining environment data, the equipment kinematic parameters, and the future predicted trajectories of the multiple dynamic obstacles; searching for a global reference path based on the spatiotemporal cost map using a path planning algorithm; and distributing the global reference path to the vehicle terminal so that the vehicle terminal controls vehicle operation based on the global reference path. In this embodiment of the invention, by introducing a visual human-computer interaction interface, users can intuitively modify and update static obstacle information in the mining environment. This effectively overcomes the shortcomings of blind spots in vehicle-side perception and the lack of efficient manual intervention methods in existing cloud control platforms. By deeply integrating manually confirmed static obstacle information with predicted trajectories of dynamic obstacles and vehicle kinematic constraints, an accurate spatiotemporal cost map is constructed in the cloud. This generates a globally optimal path that can safely avoid all known static obstacles and predict dynamic conflicts in advance. Finally, the path is sent to the vehicle via a low-latency network, greatly improving the safety, reliability, and overall efficiency of autonomous vehicles operating in complex unstructured mining environments. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of an embodiment of the obstacle-based path planning method of the present invention; Figure 2 This is a schematic diagram of another embodiment of the obstacle-based path planning method in this invention. Figure 3 This is a schematic diagram of one embodiment of the obstacle-based path planning device of the present invention; Figure 4 This is a schematic diagram of another embodiment of the obstacle-based path planning device of the present invention; Figure 5 This is a schematic diagram of one embodiment of the electronic device in this invention. Detailed Implementation

[0020] This invention provides a path planning method, apparatus, device, and storage medium based on obstacle modification. By introducing a visual human-computer interaction mechanism, it enables real-time labeling and updating of static obstacles, thereby performing path planning and effectively improving the safety and efficiency of automatic path planning in mining environments.

[0021] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0022] It is understood that the executing entity of this invention can be a path planning device based on obstacle modification, or it can be a terminal or a server; no specific limitation is made here. This embodiment of the invention will be described using a server as an example of the executing entity.

[0023] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the obstacle-modification-based path planning method in this invention includes: 101. Obtain user commands to modify static obstacle information in unstructured mining scenarios via a visual interactive interface; Unstructured mining scenarios refer to working environments lacking fixed, regular, and predictable structures. In the visual interactive interface, an interactive graphical model is built for these scenarios. Users can directly use the mouse to click and select static obstacle areas within the scene. Clicking a specific function button results in the following actions: drawing a new graphic in the selected area is considered adding a static obstacle; dragging, scaling, or otherwise altering the position or size of an existing static obstacle graphic is considered modifying it; and clicking on a static obstacle graphic and selecting the delete option is considered deleting it. These operations are captured in real-time and converted into corresponding commands for acquisition.

[0024] 102. Generate standardized static obstacle update information according to the operation instructions, and mark the priority of each static obstacle update information to obtain the static obstacle update information of each target. One static target obstacle update information corresponds to one static obstacle. Key information about static obstacles, such as obstacle type, location coordinates, and size, is extracted and integrated according to a standardized format to generate standardized static obstacle update information. A preset priority evaluation algorithm is used to calculate the priority value corresponding to each static obstacle update. This priority set algorithm comprehensively considers factors such as the impact of static obstacles on the vehicle's driving path, the order of operation commands, and the importance of the area where the static obstacle is located, calculating a priority value for each standardized static obstacle update. Based on the calculated priority values, each static obstacle update is sorted and marked with a priority level, thus obtaining the update information for each target static obstacle. Each target static obstacle update corresponds precisely to one static obstacle.

[0025] 103. Acquire mine environmental data, vehicle kinematic parameters, and predicted future trajectories of multiple dynamic obstacles; By deploying various types of sensors in the mining environment, such as lidar, to accurately detect environmental data such as mine topography and obstacle distribution, while using vehicle-mounted sensors to collect real-time kinematic parameters such as vehicle speed, acceleration, and steering angle, and then using the equipped visual camera and radar combination system to continuously track and monitor multiple dynamic obstacles, and combining advanced algorithm models to analyze and predict their motion state, the future predicted trajectory of multiple dynamic obstacles can be obtained.

[0026] 104. Based on the updated information of static obstacles for each target, mine environmental data, equipment kinematic parameters, and the future predicted trajectories of multiple dynamic obstacles, construct a spatiotemporal cost map.

[0027] For each target static obstacle update, on a pre-constructed mine base map framework, the updated static obstacle areas are marked according to their corresponding location coordinates and shape dimensions. These areas are assigned high cost values ​​to indicate that vehicles cannot pass or that passage is extremely difficult. For mine environmental data, detailed information such as terrain undulations, slopes, and potholes obtained through sensors is converted into cost factors of different levels. For example, steep slopes are assigned higher costs, and flat areas are assigned lower costs, which are then mapped to the corresponding locations on the map. Based on the vehicle's own kinematic parameters, such as maximum speed, acceleration, and turning radius, feasible and restricted areas for the vehicle under different motion states are determined. On the map, roads where the vehicle can travel safely are delineated based on these parameters. The map calculates the range of possible vehicle movement and assigns corresponding costs to areas exceeding the vehicle's movement capabilities. For example, costs are set based on the vehicle's minimum turning radius at narrow bends to ensure the map reflects the actual movement limitations of the vehicle. For the predicted trajectories of multiple dynamic obstacles, a time-slicing approach is used to divide the prediction period into multiple discrete moments. At each moment, the area occupied by the dynamic obstacle is dynamically updated on the map based on its predicted position and speed, and a cost that changes over time is set for these areas. The closer to the current moment, the higher the cost, to reflect the real-time impact of the dynamic obstacle on the vehicle's movement. By integrating all the above information, the costs from different sources are superimposed and fused to form a complete spatiotemporal cost map, in which each grid cell contains cost information in both spatial location and time dimensions.

[0028] 105. Based on the spatiotemporal cost map, a global reference path is obtained by searching using a path planning algorithm; Based on the constructed spatiotemporal cost map, an improved A* algorithm is used for global path search. Specifically, the vehicle's starting position is used as the first node in the open list, while a closed list is set up to store processed nodes. During the search process, for each node in the open list, its adjacent reachable nodes are obtained based on the spatiotemporal cost map. These adjacent nodes must avoid high-cost areas, such as high-cost blocks containing static obstacles and areas with high costs due to future trajectories of dynamic obstacles. The heuristic function value of each adjacent node is calculated. This function comprehensively considers the actual cost from the current node to the adjacent node and the estimated cost from the adjacent node to the target location, for example, using Euclidean distance or Manhattan distance combined with cost factors for estimation. Each time, the node with the smallest heuristic function value is selected from the open list as the expanded node, moved to the closed list, and its adjacent nodes are updated. If the path cost to the adjacent node through the current expanded node is lower, the parent node of the adjacent node is updated to the current expanded node, and its heuristic function value is recalculated. The above process is repeated until the expanded node reaches the target location or the open list is empty (indicating that there is no feasible path). Finally, starting from the target node, backtracking along the parent node to the starting node, the resulting path is the global reference path obtained from the search.

[0029] 106. Send the global reference path to the vehicle terminal so that the vehicle terminal can control the vehicle operation based on the global reference path.

[0030] The generated global reference path is encoded and encapsulated in a standardized data format using a stable and reliable wireless communication network, such as 5G or a dedicated vehicle-to-everything (V2X) communication link, ensuring the accuracy and integrity of data transmission before being sent to the vehicle. Upon receiving the data, the vehicle's receiving module decodes and parses it, converting the global reference path into path information that the vehicle's control system can recognize and process. Based on this path information, and combined with the vehicle's current real-time position, speed, attitude, and other state parameters, the vehicle's path tracking control algorithm calculates control commands for each actuator. By adjusting these actuators, the vehicle can follow the global reference path, thereby achieving effective control of vehicle operation based on the global reference path and ensuring the vehicle safely and efficiently completes its tasks in the complex environment of a mine.

[0031] In this embodiment of the invention, a visual human-computer interaction interface is introduced, allowing users to intuitively modify and update static obstacle information in the mining environment. This effectively overcomes the shortcomings of blind spots in vehicle-side perception and the lack of efficient manual intervention methods in existing cloud control platforms. By deeply integrating manually confirmed static obstacle information with predicted trajectories of dynamic obstacles and vehicle kinematic constraints, a precise spatiotemporal cost map is constructed in the cloud. This generates a globally optimal path that can safely avoid all known static obstacles and predict dynamic conflicts in advance. Finally, this path is transmitted to the vehicle via a low-latency network, greatly improving the safety, reliability, and overall efficiency of autonomous vehicles operating in complex, unstructured mining environments.

[0032] Please see Figure 2 Another embodiment of the obstacle-modification-based path planning method in this invention includes: 201. Obtain user commands to modify static obstacle information in unstructured mining scenarios via a visual interactive interface; It receives user commands to add new static obstacles using graphical drawing tools on a visual interactive interface. The commands include the geometry, size, and position information of the new static obstacle. It also receives user commands to modify the attributes of existing static obstacles by dragging, scaling, or rotating them. The attributes include the position, size, orientation, and type of the static obstacle. Furthermore, it receives user commands to remove a specified static obstacle by selecting and confirming a delete operation. All add, modify, and delete operations are encapsulated into a structured event data stream.

[0033] 202. Generate standardized static obstacle update information according to the operation instructions, and mark the priority of each static obstacle update information to obtain the static obstacle update information of each target. One static target obstacle update information corresponds to one static obstacle. The system parses operation instructions, identifies the operation type and corresponding static obstacle parameters. Operation types include adding, modifying, or deleting static obstacles. The parsed parameters are encapsulated into standardized static obstacle update information, which includes a unique identifier for the static obstacle, type attributes, a set of geometric parameters, and a timestamp. Based on the type attributes of the static obstacle and the importance of its location area, a corresponding priority label is assigned to each static obstacle update information to form the target static obstacle update information.

[0034] Rule-based pattern matching technology can be used to parse operation commands and identify the operation type and corresponding static obstacle parameters. Specifically, specific rule patterns are defined for add, modify, and delete operations. For add operations, when the operation command contains the behavior of generating the geometry, size, and position information of static obstacles using a graphics drawing tool, it is determined to be a add operation, and this geometry, size, and position information is extracted as the corresponding static obstacle parameters. For modify operations, if the command contains an operation identifier for dragging, scaling, rotating, or changing the type of an existing static obstacle, it is identified as a modify operation, and the modified position, size, orientation, and type information are extracted as parameters. For delete operations, when the command explicitly contains an identifier for selecting a specified static obstacle and confirming deletion, it is determined to be a delete operation. In this way, operation commands can be accurately parsed, and the operation type and corresponding static obstacle parameters can be identified.

[0035] Parameter encapsulation can be implemented using object-oriented programming. Define a static obstacle class containing attributes such as a unique identifier, type attribute, geometric parameter set, and timestamp. The unique identifier can be created using a UUID (Universally Unique Identifier) ​​generation algorithm to ensure each static obstacle has a unique identifier. The type attribute is assigned a value based on the obstacle type set by the user, such as "rock" or "mound." The geometric parameter set is stored by creating a geometric shape class, which contains attributes such as shape type, size, and position. The parsed geometric parameters are then assigned to these attributes. The timestamp is obtained by calling the system's built-in time function. Finally, using a common JSON or XML data serialization format, the unique identifier, type attribute, geometric parameter set, and timestamp are serialized according to a pre-defined format to generate standardized static obstacle update information.

[0036] A priority rule system is pre-defined, specifying the base priority value for different types of static obstacles based on their type attributes. For example, rocks and large equipment that affect vehicle safety are assigned a higher base priority, while small debris and vegetation are assigned a lower base priority. The importance of the area where the static obstacle is located is considered, dividing the area into zones of different importance. For example, critical transportation routes and hazardous work areas are designated as high-importance zones, while ordinary areas are designated as low-importance zones. For each static obstacle update, a base priority value is first determined based on its type attribute, and then adjusted according to the importance of its location area. If it is in a high-importance zone, the base priority value is increased; if it is in a low-importance zone, the base priority value remains unchanged or is appropriately reduced. Through this rule calculation, a corresponding priority tag is assigned to each static obstacle update, ultimately forming the target static obstacle update information.

[0037] 203. Acquire mine environmental data, vehicle kinematic parameters, and predicted trajectories of multiple dynamic obstacles; Multi-sensor fusion and intelligent prediction technologies can be used to acquire the required data. For mining environment data, various sensors, such as LiDAR, millimeter-wave radar, and cameras, are deployed in the mining scene. LiDAR can acquire distance and shape information of surrounding objects with high precision, millimeter-wave radar can detect the position and speed of objects, and cameras can capture rich visual information. Through sensor fusion algorithms, the data collected by these different sensors are integrated, redundancy and noise are removed, and mining environment data containing detailed information such as terrain, landforms, and obstacle distribution is generated. The kinematic parameters of the vehicle itself can be obtained through onboard sensors, such as using wheel speed sensors to measure vehicle speed, using accelerometers to obtain vehicle acceleration, using gyroscopes to determine vehicle attitude and steering angle, and then combining the vehicle's own parameters, such as wheelbase and maximum steering angle, to calculate the vehicle's kinematic parameters. For predicting the future trajectory of multiple dynamic obstacles, target detection and tracking algorithms are used. First, the position and motion state of dynamic obstacles are detected in real time by sensors such as cameras and radar. Then, machine learning models are used to predict the motion trajectory of dynamic obstacles, generating predicted trajectories of dynamic obstacles for a period of time in the future.

[0038] 204. The update information of each target static obstacle is fused into the mine environment data according to its priority and different cost decay functions to generate enhanced static environment data. Each target static obstacle is assigned a corresponding cost attenuation coefficient based on its priority. High-priority static obstacles use a slow attenuation function, while low-priority static obstacles use a fast attenuation function. The cost value of the grid containing the static obstacle and its affected area is nonlinearly superimposed according to the assigned attenuation function. The superimposed cost value is then fused with the mine environmental data to generate enhanced static environmental data.

[0039] After updating information and determining the priority for each target static obstacle, a mapping relationship between priority and attenuation coefficient is established. A threshold is set, classifying static obstacles with priorities higher than the threshold as high priority and those lower as low priority. For high-priority static obstacles, a slow attenuation function, such as an exponential attenuation function, is selected, with a large attenuation time constant, so that their cost value decreases slowly over time. For low-priority static obstacles, a fast attenuation function, such as a linear attenuation function, is used, with a small attenuation time constant, so that their cost value decreases rapidly. A grid is then divided in the mining environment to determine the grid where the static obstacle is located and its corresponding value. The affected area is defined as follows: For each grid cell, an attenuation coefficient is assigned according to the priority of the static obstacles within it. Then, the cost value at different times is calculated based on the attenuation function. These cost values ​​are then nonlinearly superimposed, for example, using a weighted summation method. The weights can be determined based on the degree of influence of the obstacles on vehicle driving. The superimposed cost value is then fused with the mine environment data obtained through multi-sensor fusion. Data fusion algorithms, such as Bayesian fusion or Kalman filter fusion, can be used to integrate the cost value information of static obstacles into the mine environment data, generating enhanced static environment data containing detailed cost information of static obstacles.

[0040] 205. Calculate the braking distance and turning radius of the vehicle based on the equipment's kinematic parameters, and convert them into safe space constraints; Based on the vehicle's maximum deceleration, current speed, and system response delay time in the equipment's kinematic parameters, the minimum safe braking distance at different speeds is calculated; based on the vehicle's wheelbase and maximum steering angle in the equipment's kinematic parameters, the minimum turning radius of the vehicle at different speeds is calculated; the minimum safe braking distance and minimum turning radius are mapped into the environmental space model to generate a safe buffer area outside the vehicle's outline, and the safe buffer area is used as a safe space constraint during path search.

[0041] For the minimum safe braking distance, calculations are performed based on vehicle kinematics formulas, using the maximum deceleration, current speed, and system reaction delay time. Specifically, the distance the vehicle travels during the system reaction delay time is first calculated (current speed multiplied by the system reaction delay time). Then, using the uniform deceleration motion formula, the distance required to decelerate from the current speed to a stop is calculated (current speed squared divided by twice the maximum deceleration). These two distances are added together to obtain the minimum safe braking distance at different speeds. For the minimum turning radius, calculations are performed based on the vehicle wheelbase and maximum steering angle using geometric relationships. The turning radius is calculated using the tangent of the wheelbase and maximum steering angle (wheelbase divided by the tangent of the maximum steering angle) to obtain the minimum turning radius at different vehicle speeds. Using the vehicle profile as a reference, a safety buffer zone is generated by extending outwards according to the calculated minimum safe braking distance and minimum turning radius.

[0042] In the path search algorithm, this safety buffer zone is used as a constraint to ensure that the searched path will not enter this zone, thereby ensuring the safety of vehicle driving.

[0043] 206. Based on enhanced static environmental data and safe space constraints, combined with the future predicted trajectories of multiple dynamic obstacles, a spatiotemporal cost map is generated; Using safety space constraints as a condition for passage feasibility, the two-dimensional environmental cost map is preprocessed to mark areas that do not conform to vehicle kinematic constraints; the future predicted trajectories of multiple dynamic obstacles are discretized in the time dimension to generate a series of dynamic obstacle occupancy grids with time steps, thus obtaining dynamic obstacle information; and the dynamic obstacle information is fused with the preprocessed two-dimensional environmental cost map in the spatiotemporal dimension to generate a three-dimensional spatiotemporal cost map.

[0044] For safety space constraints, they are used as conditions for passage feasibility. Each grid of the two-dimensional environmental cost map is traversed. Based on the safety buffer zone determined by vehicle kinematic constraints, such as minimum safe braking distance and minimum turning radius, grids that do not meet the requirements are marked. These marked grids represent areas that do not meet the vehicle kinematic constraints. For the future predicted trajectory of multiple dynamic obstacles, it is discretized in the time dimension. A time step is set, and the predicted trajectory is divided into a series of time points according to the step. At each time point, the grid position occupied by the dynamic obstacle is determined, and a series of dynamic obstacle-occupied grids with time steps are generated to obtain dynamic obstacle information. The dynamic obstacle information is then fused with the preprocessed two-dimensional environmental cost map in the spatiotemporal dimension. In the temporal dimension, the dynamic obstacle-occupied grid information of each time step is superimposed with the grid information of the corresponding time point in the two-dimensional environmental cost map. In the spatial dimension, the grids occupied by the dynamic obstacle are marked on the environmental cost map. At the same time, the impact of the movement of the dynamic obstacle on the surrounding area is considered, and the cost value of the corresponding grid is adjusted. Finally, a three-dimensional spatiotemporal cost map is generated, which contains both static environmental information and dynamic obstacle information.

[0045] 207. Based on the spatiotemporal cost map, a global reference path is obtained by searching using a path planning algorithm; The spatiotemporal cost map is converted into a graph structure, where grids correspond to nodes, and adjacency relationships between grids correspond to edges between nodes. The weight of an edge is determined by the cost value of the grid; the higher the cost value, the greater the weight. A suitable graph search algorithm is selected, such as the A* algorithm. The A* algorithm guides the search direction through an evaluation function, which typically consists of two parts: the actual cost from the starting point to the current node and a heuristically estimated cost from the current node to the destination. During the search, the algorithm starts from the starting point and gradually expands the nodes, selecting the node with the smallest evaluation function value for expansion each time, until the destination is found. When expanding a node, it checks whether the node satisfies the safe space constraint, i.e., whether it is outside the marked area that does not conform to vehicle kinematic constraints. It also considers the impact of dynamic obstacles occupying grids and avoids areas where dynamic obstacles may appear. By continuously expanding and evaluating nodes, a global reference path from the starting point to the destination is finally found.

[0046] 208. Send the global reference path to the vehicle terminal so that the vehicle terminal can control the vehicle operation based on the global reference path.

[0047] A transmission method based on 5G network slicing and edge computing can be adopted. Specifically, a dedicated low-latency network slice is created in the 5G core network. By configuring network slicing policies, sufficient network resources, such as bandwidth and latency guarantees, are allocated to this slice to ensure low latency and high reliability of data transmission. Edge computing nodes are deployed in scenarios such as mines, and the calculation and transmission tasks of the global reference path are partially offloaded to the edge computing nodes. After the global reference path is generated, the path data is transmitted to the edge computing nodes, which preprocess and optimize the path data, such as compressing data size and extracting key information, to reduce the amount of data transmitted. Then, the edge computing nodes send the processed global reference path data to the vehicle through low-latency network slices. During the sending process, a reliable data transmission protocol, such as TCP, is used to ensure the integrity and correctness of the data. At the same time, the high speed and low latency characteristics of the 5G network are used to monitor the data transmission status in real time. If a transmission anomaly occurs, retransmission or adjustment of the transmission strategy is performed in a timely manner. After receiving the global reference path, the vehicle controls the vehicle according to the path information, such as adjusting the vehicle speed and steering angle, to achieve operation based on the global reference path.

[0048] In this embodiment of the invention, manual intervention is combined with automated planning. Through a visual interactive interface, operators can directly modify and annotate static obstacle information in the mining environment, effectively compensating for the limitations of vehicle-side perception. By establishing a standardized obstacle update information processing flow and realizing dynamic environment modeling based on a priority mechanism, and then integrating multi-dynamic obstacle prediction trajectories and vehicle kinematic constraints, a spatiotemporal cost map that accurately reflects the actual working environment is constructed. The global reference path generated based on this map not only ensures effective avoidance of static obstacles but also achieves proactive avoidance of dynamic obstacles. Finally, the optimized path is sent to the vehicle in real time through a vehicle-cloud collaboration mechanism, thereby comprehensively improving the operational safety, path planning accuracy, and system operating efficiency of autonomous vehicles in complex unstructured mining environments.

[0049] The path planning method based on obstacle modification in the embodiments of the present invention has been described above. The path planning device based on obstacle modification in the embodiments of the present invention will be described below. Please refer to [link / reference]. Figure 3 One embodiment of the obstacle-modification-based path planning device of the present invention includes: The first acquisition module 301 is used to acquire the user's operation instructions on the visual interactive interface to modify the static obstacle information in the unstructured scene of the mine. The processing module 302 is used to generate standardized static obstacle update information according to the operation instructions, and to mark the priority of each static obstacle update information to obtain the target static obstacle update information. One target static obstacle update information corresponds to one obstacle. The second acquisition module 303 is used to acquire mine environmental data and the equipment kinematic parameters of the vehicle itself; Module 304 is used to fuse information on the updated static target obstacles, mine environment data and equipment kinematic parameters, and to predict the future trajectories of multiple dynamic obstacles in order to construct a spatiotemporal cost map. Search module 305 is used to search based on the spatiotemporal cost map using a path planning algorithm to obtain a global reference path; The distribution module 306 is used to distribute the global reference path to the vehicle terminal so that the vehicle terminal can control the vehicle operation based on the global reference path.

[0050] In this embodiment of the invention, by introducing a visual human-computer interaction interface, users can intuitively modify and update static obstacle information in the mining environment. This effectively overcomes the shortcomings of blind spots in vehicle-side perception and the lack of efficient manual intervention methods in existing cloud control platforms. By deeply integrating manually confirmed static obstacle information with predicted trajectories of dynamic obstacles and vehicle kinematic constraints, an accurate spatiotemporal cost map is constructed in the cloud. This generates a globally optimal path that can safely avoid all known static obstacles and predict dynamic conflicts in advance. Finally, the path is sent to the vehicle via a low-latency network, greatly improving the safety, reliability, and overall efficiency of autonomous vehicles operating in complex unstructured mining environments.

[0051] Please see Figure 4 Another embodiment of the obstacle-modification-based path planning device in this invention includes: The first acquisition module 301 is used to acquire the user's operation instructions on the visual interactive interface to modify the static obstacle information in the unstructured scene of the mine. The processing module 302 is used to generate standardized static obstacle update information according to the operation instructions, and to mark the priority of each static obstacle update information to obtain the target static obstacle update information. One target static obstacle update information corresponds to one obstacle. The second acquisition module 303 is used to acquire mine environmental data and the equipment kinematic parameters of the vehicle itself; Module 304 is used to fuse information on static obstacles, mine environment data and equipment kinematic parameters, and to predict the future trajectories of multiple dynamic obstacles in order to construct a spatiotemporal cost map. Search module 305 is used to search based on the spatiotemporal cost map using a path planning algorithm to obtain a global reference path; The distribution module 306 is used to distribute the global reference path to the vehicle terminal so that the vehicle terminal can control the vehicle operation based on the global reference path.

[0052] Optionally, the first acquisition module 301 can be specifically used for: It receives user commands to add new static obstacles using graphical drawing tools on a visual interactive interface. The commands include the geometry, size, and position information of the new static obstacle. It also receives user commands to modify the attributes of existing static obstacles by dragging, scaling, or rotating them. The attributes include the position, size, orientation, and type of the static obstacle. Furthermore, it receives user commands to remove a specified static obstacle by selecting and confirming a delete operation. All add, modify, and delete operations are encapsulated into a structured event data stream.

[0053] Optionally, the processing module 302 can be specifically used for: The system parses operation instructions, identifies the operation type and corresponding static obstacle parameters. Operation types include adding, modifying, or deleting static obstacles. The parsed parameters are encapsulated into standardized static obstacle update information, which includes a unique identifier for the static obstacle, type attributes, a set of geometric parameters, and a timestamp. Based on the type attributes of the static obstacle and the importance of its location area, a corresponding priority label is assigned to each static obstacle update information to form the target static obstacle update information.

[0054] Optionally, building module 304 includes: The first generation unit 3041 is used to integrate the update information of each target static obstacle into the mine environment data according to its priority and with different cost decay functions to generate enhanced static environment data. The processing unit 3042 is used to calculate the braking distance and turning radius of the vehicle based on the kinematic parameters of the equipment, and convert them into safety space constraints; The second generation unit 3043 is used to generate a spatiotemporal cost map based on enhanced static environmental data and safe space constraints, combined with the future predicted trajectories of multiple dynamic obstacles.

[0055] Optionally, the first generating unit 3041 can be specifically used for: Each target static obstacle is assigned a corresponding cost attenuation coefficient based on its priority. High-priority static obstacles use a slow attenuation function, while low-priority static obstacles use a fast attenuation function. The cost value of the grid containing the static obstacle and its affected area is nonlinearly superimposed according to the assigned attenuation function. The superimposed cost value is then fused with the mine environmental data to generate enhanced static environmental data.

[0056] Optionally, the processing unit 3042 may be specifically used for: Based on the vehicle's maximum deceleration, current speed, and system response delay time in the equipment's kinematic parameters, the minimum safe braking distance at different speeds is calculated; based on the vehicle's wheelbase and maximum steering angle in the equipment's kinematic parameters, the minimum turning radius of the vehicle at different speeds is calculated; the minimum safe braking distance and minimum turning radius are mapped into the environmental space model to generate a safe buffer area outside the vehicle's outline, and the safe buffer area is used as a safe space constraint during path search.

[0057] Optionally, the second generation unit 3043 can be specifically used for: Based on enhanced static environmental data, an initial two-dimensional environmental cost map is constructed. The safe space constraint is used as a condition for passage feasibility. The two-dimensional environmental cost map is preprocessed to mark areas that do not meet vehicle kinematic constraints. The future predicted trajectories of multiple dynamic obstacles are discretized in the time dimension to generate a series of dynamic obstacle occupancy grids with time steps, thus obtaining dynamic obstacle information. The dynamic obstacle information is then fused with the preprocessed two-dimensional environmental cost map in the spatiotemporal dimension to generate a three-dimensional spatiotemporal cost map.

[0058] In this embodiment of the invention, manual intervention is combined with automated planning. Through a visual interactive interface, operators can directly modify and annotate static obstacle information in the mining environment, effectively compensating for the limitations of vehicle-side perception. By establishing a standardized obstacle update information processing flow and realizing dynamic environment modeling based on a priority mechanism, and then integrating multi-dynamic obstacle prediction trajectories and vehicle kinematic constraints, a spatiotemporal cost map that accurately reflects the actual working environment is constructed. The global reference path generated based on this map not only ensures effective avoidance of static obstacles but also achieves proactive avoidance of dynamic obstacles. Finally, the optimized path is sent to the vehicle in real time through a vehicle-cloud collaboration mechanism, thereby comprehensively improving the operational safety, path planning accuracy, and system operating efficiency of autonomous vehicles in complex unstructured mining environments.

[0059] above Figure 3 and Figure 4 The obstacle-modification-based path planning device in this embodiment of the invention will be described in detail from the perspective of modular functional entities. The electronic device in this embodiment of the invention will be described in detail from the perspective of hardware processing.

[0060] See Figure 5 As shown, the electronic device includes a processor 500 and a memory 501. The memory 501 stores machine-executable instructions that can be executed by the processor 500. The processor 500 executes the machine-executable instructions to implement the above-described obstacle-modification-based path planning method.

[0061] Furthermore, Figure 5The electronic device shown also includes a bus 502 and a communication interface 503. The processor 500, the communication interface 503 and the memory 501 are connected via the bus 502.

[0062] The memory 501 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 503 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc. The bus 502 may be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 5 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0063] The processor 500 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 500 or by instructions in software form. The processor 500 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this disclosure. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this disclosure can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules may reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 501. The processor 500 reads the information in memory 501 and, in conjunction with its hardware, completes the method steps of the aforementioned embodiment.

[0064] The present invention also provides an electronic device, the computer device including a memory and a processor, wherein the memory stores computer-readable instructions, and when executed by the processor, the computer-readable instructions cause the processor to perform the steps of the obstacle-modification-based path planning method described in the above embodiments. The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions, and when the instructions are executed on a computer, the computer causes the computer to perform the steps of the obstacle-modification-based path planning method.

[0065] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0066] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0067] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention 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 of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A path planning method based on obstacle modification, characterized in that, The obstacle-modification-based vehicle-cloud cooperative planning method includes: Obtain user commands to modify static obstacle information in unstructured mining scenarios via a visual interactive interface; Standardized static obstacle update information is generated according to the operation instructions, and priority is marked for each static obstacle update information to obtain the static obstacle update information for each target. One static target obstacle update information corresponds to one static obstacle. Acquire mine environmental data, vehicle kinematic parameters, and predicted future trajectories of multiple dynamic obstacles; Based on the updated information of each target static obstacle, the mine environment data, the equipment kinematic parameters, and the future predicted trajectories of the multiple dynamic obstacles, a spatiotemporal cost map is constructed. Based on the aforementioned spatiotemporal cost map, a global reference path is obtained through a path planning algorithm. The global reference path is sent to the vehicle terminal so that the vehicle terminal controls the vehicle operation based on the global reference path.

2. The path planning method based on obstacle modification according to claim 1, characterized in that, The process of obtaining user operation instructions includes: The system receives an operation command from a user on the visual interactive interface to add a new static obstacle using a graphical drawing tool. The command includes the geometric shape, size, and position information of the new static obstacle. Receive instructions from the user to modify the attributes of an existing static obstacle through drag, zoom, or rotate operations. The attributes include the static obstacle's position, size, orientation, and type. Receives instructions from the user to remove a specified static obstacle by selecting and confirming the deletion operation; Add, modify, and delete operations are uniformly encapsulated into a structured event data stream.

3. The path planning method based on obstacle modification according to claim 1, characterized in that, The step of generating standardized obstacle update information according to the operation instructions and prioritizing each obstacle update information to obtain the target obstacle update information includes: The operation instructions are parsed to identify the operation type and the corresponding static obstacle parameters. The operation type includes adding, modifying, or deleting static obstacles. The parsed parameters are encapsulated into static obstacle update information in a standardized format, which includes a unique static obstacle identifier, type attribute, set of geometric parameters, and timestamp. Based on the type attributes of static obstacles and the importance of their location areas, a corresponding priority label is assigned to each static obstacle update information to form the target static obstacle update information.

4. The path planning method based on obstacle modification according to claim 1, characterized in that, The construction of a spatiotemporal cost map based on the updated information of each target static obstacle, the mine environment data, the equipment kinematic parameters, and the future predicted trajectories of the multiple dynamic obstacles includes: The updated information of each target static obstacle is fused into the mine environment data according to its priority and different cost decay functions to generate enhanced static environment data. The braking distance and turning radius of the vehicle are calculated based on the kinematic parameters of the device, and then converted into safety space constraints. Based on the enhanced static environmental data and the safety space constraints, and combined with the future predicted trajectories of the multiple dynamic obstacles, a spatiotemporal cost map is generated.

5. The path planning method based on obstacle modification according to claim 4, characterized in that, The step of fusing the update information of each target obstacle into the mine environment data according to its priority and with different cost decay functions to generate enhanced static environment data includes: Each target static obstacle is assigned a corresponding cost decay coefficient based on its priority. High-priority static obstacles use a slow decay function, while low-priority static obstacles use a fast decay function. The cost of the grid containing the static obstacle and its affected area is calculated nonlinearly by superimposing the values ​​according to the assigned attenuation function. The superimposed cost value is then fused with the mine environmental data to generate enhanced static environmental data.

6. The path planning method based on obstacle modification according to claim 4, characterized in that, The calculation of the vehicle's braking distance and turning radius based on the device's kinematic parameters, and the conversion of these into safety space constraints, includes: Based on the vehicle's maximum deceleration, current speed, and system reaction delay time in the device's kinematic parameters, calculate the minimum safe braking distance at different speeds; Based on the vehicle's wheelbase and maximum steering angle in the device's kinematic parameters, calculate the vehicle's minimum turning radius at different speeds; The minimum safe braking distance and minimum turning radius are mapped onto the environmental space model to generate a safe buffer area outside the vehicle outline. This safe buffer area is then used as a safe space constraint during path search.

7. The path planning method based on obstacle modification according to claim 4, characterized in that, The generation of a spatiotemporal cost map based on the enhanced static environmental data and the safety space constraints, combined with the predicted future trajectories of the multiple dynamic obstacles, includes: Based on the enhanced static environmental data, an initial two-dimensional environmental cost map is constructed; Using the safety space constraint as a feasibility condition, the two-dimensional environmental cost map is preprocessed to mark areas that do not conform to vehicle kinematic constraints; The predicted future trajectories of the multiple dynamic obstacles are discretized in the time dimension to generate a series of dynamic obstacle-occupying grids with time steps, thereby obtaining dynamic obstacle information. The dynamic obstacle information is then fused with the preprocessed two-dimensional environmental cost map in the spatiotemporal dimension to generate a three-dimensional spatiotemporal cost map.

8. A path planning device based on obstacle modification, characterized in that, The obstacle-modification-based path planning device includes: The first acquisition module is used to acquire user operation instructions on the visual interactive interface to modify static obstacle information in unstructured mining scenarios; The processing module is used to generate standardized static obstacle update information according to the operation instructions, and to mark the priority of each static obstacle update information to obtain the target static obstacle update information. One target static obstacle update information corresponds to one obstacle. The second acquisition module is used to acquire mine environmental data and the vehicle's own equipment kinematic parameters. The construction module is used to fuse the updated information of each target static obstacle, the mine environment data and the kinematic parameters of the equipment, and to predict the future trajectory of multiple dynamic obstacles in order to construct a spatiotemporal cost map. The search module is used to search based on the spatiotemporal cost map using a path planning algorithm to obtain a global reference path; The distribution module is used to distribute the global reference path to the vehicle terminal, so that the vehicle terminal can control the vehicle operation based on the global reference path.

9. An electronic device, characterized in that, The electronic device includes: a memory and at least one processor, wherein the memory stores instructions; The at least one processor invokes the instructions in the memory to cause the electronic device to execute the obstacle-modification-based path planning method as described in any one of claims 1-7.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement the obstacle-based path planning method as described in any one of claims 1-7.

Images