A method and system for planning uniform loading and unloading paths for multiple vehicles in a mine dump

Abstract

The present invention discloses a method and system for planning paths for uniform loading and unloading of multiple vehicles in a mine dump, comprising: step 1, after entering the planning starting point, obtaining status information, trajectory information, and visit count information of other vehicles in the work area; step 2, initializing a potential field map based on map data and adjusting the potential field of the potential field map according to the paths of other vehicles in the work area; step 3, generating a work path and an exit path using a planning method using a variable potential field; step 4, after the work is completed, returning to step 2; step 3 includes: step 31, planning the work path; and step 32, planning the exit path. The present invention has high real-time performance, security, and practical applicability.

Description

Technical Field

[0001] The present invention relates to the field of mine automatic driving technology, and in particular to a method and system for planning paths for uniform loading and unloading of multiple vehicles in a mine dump. Background Art

[0002] Multi-vehicle trajectory planning for unstructured roads is a key technology for improving vehicle traffic efficiency and reducing trajectory conflicts in scenarios such as mines and parking lots. Existing research focuses primarily on multi-vehicle trajectory planning for drones and robots, with limited research on four-wheeled vehicles. Furthermore, most research on multi-vehicle trajectory planning for four-wheeled vehicles has been conducted in structured environments. For example, extensive research has been conducted on multi-vehicle trajectory planning for structured roads, primarily focusing on priority-based, reaction-based, numerical optimization-based, and speed coordination-based planning methods. There are three main approaches to multi-vehicle trajectory planning for unstructured roads. One employs the RRT* algorithm to simultaneously plan multiple vehicle trajectories in the time domain and then expands these nodes using Dubins curves. The second approach involves designing a heuristic function that dynamically prioritizes vehicles based on the number of their homology classes to reduce the probability of deadlock. The three-dimensional A* algorithm is then used to search for individual vehicle trajectories in space and time. The third method designs a failure-resistant hybrid A*. It first searches for an initial path, then designs a circular risk zone to scale the collision avoidance constraints, constructs an intermediate optimization problem to reduce the computational dimension, and obtains the optimal solution that satisfies all constraints through continuous iteration.

[0003] Whether on structured or unstructured roads, as the number of vehicles increases, the collision avoidance constraints for each vehicle increase exponentially, often leading to longer calculation times. Using separate trajectory planning methods for each vehicle can lead to trajectories being too close together, triggering a safe stop for the intelligent vehicle and increasing the probability of deadlock. Furthermore, for special unstructured areas like work zones, it's also important to consider uniformity of operations, avoiding repeated operations at the same work point or allowing a single work point to remain idle for extended periods of time to ensure continuous loading and unloading.

[0004] In summary, the current multi-vehicle planning method has low real-time performance and poor stability, and cannot meet the needs of actual engineering applications.

[0005] For example, the method provided by the prior art is: first calculate the time taken by each vehicle to search for a path using RRT* without considering collisions, and divide the travel time of each vehicle by the maximum travel time in the fleet to ensure that all vehicles will arrive at the destination at the same time. In each expansion step, the random points of all vehicles are expanded, and then connected using the Dubins curve. Finally, the expanded path is checked for collisions. If there is a collision, it will be replanned until the path expanded in this step is safe and effective. Repeat the above expansion process until each vehicle reaches the destination. However, when using RRT* for expansion, many invalid nodes will appear, and these invalid nodes will affect each other. When the map area is small or the number of vehicles is large, processing these invalid nodes will waste a lot of computing power.

[0006] For example, the method provided by the second prior art is to calculate the number of different paths for vehicles to reach the destination based on algebraic topology theory, dynamically calculate the number of homology classes in the entire time domain, and the fewer homology classes, the higher the priority. The high-priority and low-priority sets of each vehicle in each step are always maintained. Whenever the trajectory of a vehicle in the high-priority set changes, the three-dimensional A* algorithm is re-used to plan the trajectory for the vehicle. However, dynamic priority sorting will cause the priorities of each vehicle to change frequently. This unstable decision-making ultimately leads to frequent switching of the speeds of each vehicle, wasting a lot of energy and failing to meet energy-saving requirements. The trajectories of each vehicle are closely connected, and when vehicle planning fails or control errors occur, it is difficult to handle.

[0007] For example, the method provided by prior art three first uses a fail-safe hybrid A* algorithm to search for an initial path, assigning an optimal speed to the path to form an initial trajectory. The method then calculates whether the distance between each vehicle's trajectory points is within the circular ring of the current vehicle. If so, the collision avoidance constraints for both vehicles are incorporated into the optimal control problem; otherwise, the collision avoidance constraints between them are ignored. The optimal control problem is then adjusted by scaling the upper and lower bounds of the ring, and an iterative approach is used to find an approximate feasible trajectory where the collision degree of each vehicle is within a certain range. Finally, the lower bound is fixed at zero, forming a circular risk zone around the vehicle. The radius of the risk circle is continuously increased, and iterations are repeated until the optimal solution for multiple vehicles satisfies all constraints. However, while setting an intermediate optimal control problem and solving it iteratively can increase the success rate of solving the optimal control problem to a certain extent, it is difficult to ensure a fast solution speed. As the number of vehicles increases, the dimension of the collision avoidance constraints in the optimal control problem increases exponentially, often making it impossible to successfully solve the problem. This makes it difficult to meet real-time requirements at intersections and in areas with high traffic flow. Summary of the Invention

[0008] The purpose of the present invention is to provide a method and system for planning paths for uniform loading and unloading of multiple vehicles in a mine dump, which has high real-time performance, safety and practical applicability.

[0009] To achieve the above objectives, the present invention provides a method for planning paths for uniform loading and unloading of multiple vehicles in a mine dump, which comprises:

[0010] Step 1: After entering the planned starting point, the vehicle obtains the status information, trajectory information and visit count of other vehicles in the operation area from the background;

[0011] Step 2: Initialize the potential field map based on the map data and adjust the potential field of the potential field map according to the paths of other vehicles in the work area;

[0012] Step 3: Generate the operation path and exit path through the planning method of the variable potential field;

[0013] Step 4: After the job is completed, return to step 2;

[0014] Among them, step 3 specifically includes:

[0015] If it is a planned operation path, go to step 31, otherwise go to step 32;

[0016] Step 31, planning the work path: vehicles in the work area are divided into two categories: entering vehicles and leaving vehicles. The paths of the entering vehicles are located in the potential field map output in step 2. Potential fields are pre-set on both sides of the paths of the entering vehicles. The potential fields generated by different vehicles are linearly superimposed. A path search is performed from the entry point to the work point to generate multiple candidate paths. From the output candidate paths, the optimal path is selected using a uniform evaluation function as the work path.

[0017] Step 32, planning the exit path: Clear the potential field generated by the incoming vehicle during the operation path planning. At the same time, based on the status information of other vehicles provided by the background, locate the path of the outgoing vehicle in the potential field map output in step 2, and pre-set potential fields on both sides of its path. Search for a path from the operation point to the exit point to generate the exit path.

[0018] Furthermore, the method of presetting the potential field in step 31 includes:

[0019] When d car <d default When the potential field is g car [1-(d car / d default )], when d car ≥d default When , the potential field size is 0;

[0020] Where g car The potential field value generated by the set vehicle path satisfies g car ≥g default, which can ensure that the impact of the vehicle potential field is greater than the impact of the initialization potential field, d car is the lateral offset of the projection of the point to the vehicle path, d default The influence distance of the setting.

[0021] Furthermore, the method of presetting the potential field in step 32 includes:

[0022] When planning the exit route, an additional potential field is applied in the entry route concentration area to prevent incoming vehicles from occupying the exit position or outgoing vehicles from occupying the entry position.

[0023] Furthermore, step 2 specifically includes:

[0024] First find the farthest point P from the boundary in the working area c , P c The distance from the boundary is set to d k ; Then find the distance d from the boundary in the working area k -β point set: {P 11 ,P 12 ,....,P 1m}, then find the distance d from the boundary of the working area in the working area. k -2β point set: {P 21 ,P 22 ,....,P 2m}, repeat this process to find the distance d from the boundary in the working area k -nβ point set {P n1 ,P n2 ,....,P nm} point set; where β is the potential field attenuation distance and can be optimized according to the area of ​​the working area, m is the number of discrete points on the potential field boundary and is determined by the point set spacing, n is the number of potential field layers and is determined by the area of ​​the working area, and the regional center potential field of the annular artificial potential field is g default , based on the distance point P c The distance decreases in turn, and the potential field value decreases to 0 in the blank position around the center of the area. The vehicle needs to maintain a collision avoidance distance with the boundary of the circular artificial potential field. The potential field near the boundary is set to g obs .

[0025] Furthermore, step 4 specifically includes:

[0026] Step 41: Sample m on the i-th candidate path with a spacing of Δs i path points, and the potential field values ​​of all path points are accumulated to obtain the total potential field value of the path. If the path planning fails, Set to positive infinity, and calculate the evaluation value of the i-th candidate path through the following evaluation function

[0027]

[0028] Where, The additional potential field for the number of visits to the kth operating point is used to ensure that the number of visits to each operating point tends to be consistent. It is the main influencing indicator of uniform operation. The time-attenuated additional potential field for the kth operating point is used to adjust the working interval of the operating point to prevent vehicles from waiting near the operating point and idling for a long time;

[0029] Step 42, select The smallest path is taken as the optimal path.

[0030] The present invention also provides a multi-vehicle uniform loading and unloading path planning system for a mine dump, which comprises:

[0031] The vehicle information acquisition module is used to obtain the status information, trajectory information and visit count of other vehicles in the operation area from the background after the vehicle enters the planned starting point;

[0032] The planning module is used to initialize the potential field map based on the map data and adjust the potential field of the potential field map according to the paths of other vehicles in the work area;

[0033] A path planning module, which is used to generate an operation path and an exit path through a planning method of a variable potential field;

[0034] The job completion judgment module is used to continue the potential field adjustment by the planning module after the job is completed;

[0035] Among them, the path planning module includes:

[0036] The operation path planning unit is used to classify vehicles in the operation area into two categories: entering vehicles and leaving vehicles. The paths of the entering vehicles are located in the potential field map output in step 2. Potential fields are pre-set on both sides of the paths of the entering vehicles. The potential fields generated by different vehicles are linearly superimposed. A path search is performed from the entry point to the operation point to generate multiple candidate paths. From the output candidate paths, the optimal path is selected as the operation path through the homogenization evaluation function.

[0037] The exit path planning unit is used to clear the potential field generated by the incoming vehicle during the operation path planning. At the same time, based on the status information of other vehicles provided by the background, the path of the outgoing vehicle is located in the potential field map output by the planning module, and potential fields are pre-set on both sides of its path. The path is searched from the operation point to the exit point to generate the exit path.

[0038] Furthermore, the method of presetting the potential field in the operation path planning unit includes:

[0039] When d car<d default When the potential field is g car [1-(d car / d default )], when d car ≥d default When , the potential field size is 0;

[0040] Where g car The potential field value generated by the set vehicle path satisfies g car ≥g default , which can ensure that the impact of the vehicle potential field is greater than the impact of the initialization potential field, d car is the lateral offset of the projection of the point to the vehicle path, d default The influence distance of the setting.

[0041] Furthermore, the method of presetting the potential field in the exit path planning unit includes:

[0042] When planning the exit route, an additional potential field is applied in the entry route concentration area to prevent incoming vehicles from occupying the exit position or outgoing vehicles from occupying the entry position.

[0043] Furthermore, the planning module specifically includes:

[0044] First find the farthest point P from the boundary in the working area c , P c The distance from the boundary is set to d k ; Then find the distance d from the boundary in the working area k -β point set: {P 11 ,P 12 ,....,P 1m}, then find the distance d from the boundary of the working area in the working area. k -2β point set: {P 21 ,P 22 ,....,P 2m}, repeat this process to find the distance d from the boundary in the working area k -nβ point set {P n1 ,P n2 ,....,P nm} point set; where β is the potential field attenuation distance and can be optimized according to the area of ​​the working area, m is the number of discrete points on the potential field boundary and is determined by the point set spacing, n is the number of potential field layers and is determined by the area of ​​the working area, and the regional center potential field of the annular artificial potential field is g default , based on the distance point P c The distance decreases in turn, and the potential field value decreases to 0 in the blank position around the center of the area. The vehicle needs to maintain a collision avoidance distance with the boundary of the circular artificial potential field. The potential field near the boundary is set to g obs .

[0045] Furthermore, the path planning module specifically includes:

[0046] Sample m on the i-th candidate path with a spacing of Δs i path points, and the potential field values ​​of all path points are accumulated to obtain the total potential field value of the path. If the path planning fails, Set to positive infinity, and calculate the evaluation value of the i-th candidate path through the following evaluation function choose The smallest path is taken as the optimal path;

[0047]

[0048] Where, The additional potential field for the number of visits to the kth operating point is used to ensure that the number of visits to each operating point tends to be consistent. It is the main influencing indicator of uniform operation. The time-attenuated additional potential field for the kth operating point is used to adjust the working interval of the operating point and prevent vehicles from waiting near the operating point and idling for a long time.

[0049] The present invention first initializes a default potential field, ensuring orderly vehicle entry and exit from the work area. Furthermore, when planning routes, the potential fields of existing routes are considered to reduce spatial intersections between routes. High potential fields are set in reverse sections to minimize deadlocks. Finally, a cumulative potential field is set based on the number of visits to the work point to ensure uniform loading and unloading.

[0050] Compared to the method in prior art 1 in the background art, the present invention takes into account reversing conditions and vehicle kinematics. Compared to prior art 2 in the background art, the present invention designs a planning method with less spatial interference, less intense competition for right of way between vehicles, more orderly operation routes, and a lower likelihood of deadlock. Compared to prior art 3 in the background art, the present invention requires significantly less online computation and achieves a significant improvement in computational speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A schematic diagram of a work area provided in an embodiment of the present invention.

[0052] Figure 2 Schematic diagram of work area planning provided by an embodiment of the present invention

[0053] Figure 3 This is a framework diagram of a multi-vehicle uniform loading and unloading path planning system for a mine dump provided by an embodiment of the present invention.

[0054] Figure 4 A schematic diagram of a central potential field diagram provided by an embodiment of the present invention.

[0055] Figure 5 A schematic diagram of a potential field planning provided by an embodiment of the present invention.

[0056] Figure 6 A schematic diagram of work area planning provided in an embodiment of the present invention.

[0057] Figure 7 A candidate path planning graph provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0058] In the accompanying drawings, the same or similar reference numerals are used to represent the same or similar elements or elements with the same or similar functions. The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0059] In the description of the present invention, the terms "center", "longitudinal", "lateral", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like to indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction. Therefore, they should not be understood as limiting the scope of protection of the present invention.

[0060] In unstructured roads, loading and unloading areas are extracted as follows: Figure 1 As shown, the irregular driving area with clear boundaries includes the entry point P of the vehicle into the area. in , the exit point P of the vehicle leaving the area out , and x operating points P1, P2, P3....P distributed along the boundary to complete the task in the area x .by Figure 1 For example, the vehicle will enter from the entry point P in Enter the loading and unloading area, complete the loading and unloading tasks at any operation point P1, P2, P3, and exit from the exit point P out Driving out. Loading and unloading include unloading and loading, so the operation area includes the loading area and the unloading area. The present invention is applicable to both areas. The operation points in the two areas correspond to the loading point and the operation point, respectively.

[0061] In an unmanned operation system, after a vehicle enters the operation area, the backend assigns it an operation task (loading and unloading), and the vehicle-side planning module plans a trajectory to the operation point. The trajectory of existing vehicles in the operation area will affect the planning of subsequent vehicle trajectories. The area entrances and exits and the operation points are the main conflict areas. For example, using traditional hybrid A* planning, such as Figure 2As shown, the blue path of vehicle 1 and the yellow path of vehicle 2 overlap for a long time before reaching their respective work points. At the same time, the reversing section of vehicle 2 conflicts with the path of vehicle 1 to a certain extent. As the number of vehicles increases, the degree of these conflicts will become more severe.

[0062] In summary, multi-vehicle planning in work areas presents the following three issues: First, the shortest path planning objective leads to congestion in the central area. Second, the interference between reverse and forward sections can easily cause serious path conflicts among multiple vehicles, and speed planning cannot resolve these conflicts, leading to a deadlock. Third, the random selection of work points can easily lead to uneven expansion, such as long-term idleness or unused work points, which requires manual intervention to resolve.

[0063] In order to solve the above problems, the embodiment of the present invention proposes a method for planning paths for uniform loading and unloading of multiple vehicles in a mine dump, including: Figure 3 As shown:

[0064] Step 1: After entering the planned starting point, the vehicle obtains the status information, trajectory information, and visit count information of other vehicles in the work area from the background. Among them, other vehicles can be understood as other vehicles with work tasks.

[0065] Step 2: Initialize the potential field map based on the map data and adjust the potential field of the potential field map according to the paths of other vehicles in the work area.

[0066] Step 3: Generate the work path and exit path using a planning method based on a variable potential field. In this embodiment, the planning method based on a variable potential field can be improved upon by using a hybrid A* algorithm, expanding it based on the potential field value, and sequentially generating multiple human-like paths with each work point as the planning endpoint.

[0067] Step 4: After the operation is completed, return to step 2, the planning module reinitializes the potential field map, adjusts the potential field again based on the path of other vehicles, and plans the vehicle's exit path.

[0068] In one embodiment, based on observations of human driving behavior, it is apparent that when selecting a driving path in a work area, drivers tend to choose a path that is closer to the boundary of the area. Figure 4 As shown, the driving paths tend to be distributed around the center of the area. This driving behavior greatly reduces the occurrence of overlap on the road, thereby minimizing the degree of conflict between vehicles. By adding a circular artificial potential field in the operation area, the unmanned vehicle is guided to simulate the driving behavior of a human driver. Therefore, step 2 specifically includes:

[0069] First find the farthest point P from the boundary in the working area c , P c The distance from the boundary is set to d k ; Then find the distance d from the boundary in the working areak -β point set: {P 11 ,P 12 ,....,P 1m}, then find the distance d from the boundary of the working area in the working area. k -2β point set: {P 21 ,P 22 ,....,P 2m}, repeat this process to find the distance d from the boundary in the working area k -nβ point set {P n1 ,P n2 ,....,P nm} point set.

[0070] Where β is the potential field attenuation distance and can be optimized according to the area of ​​the work area. It can be set reasonably according to the size of the area. The larger the setting, the faster the calculation speed, and the smaller the setting, the higher the calculation accuracy. It can usually be set to 10m. m is the number of discrete points on the potential field boundary and is determined by the point set spacing. It is usually set to 0.1m. n is the number of potential field layers and is determined by the area of ​​the work area. For example: S≈π*(nβ) 2 .like Figure 4 As shown, the closer to the center area, the stronger the potential field. The potential field in the center of the annular artificial potential field is g default , usually set to 1, based on the distance point P c The distance between the potential field and the area decreases in turn, and the potential field value decreases to 0 in the blank area around the center of the area. In addition, the vehicle needs to maintain a collision avoidance distance from the boundary of the circular artificial potential field, and the potential field near the boundary is set to g obs , g obs Usually set to g default 10 times of that.

[0071] The embodiment of the present invention is based on the original hybrid A* algorithm. When calculating the cost function G value, it not only considers the turning cost and the distance cost, but also the potential field cost. After this improvement, the hybrid A* algorithm will be affected by the existing potential field value during the expansion process and expand in the desired direction. Figure 4 As shown, under the action of the potential field, vehicle A forms a human-like operating path.

[0072] When planning under the same potential field, the same starting point will produce the same path. In the work area, although there are multiple endpoints to choose from, the same potential field can still cause long overlapping sections in the middle of each vehicle's path. To address this issue, the present invention provides a planning method with a variable potential field. Path planning for vehicles entering and exiting the work area is divided into two parts: the first part is work path planning, and the second part is exit path planning.

[0073] In one embodiment, step 3 specifically includes:

[0074] If the operation path is to be planned, the process proceeds to step 31 ; otherwise, the process proceeds to step 32 .

[0075] Step 31, planning the operation path: The scheduling backend divides the vehicles in the operation area into two categories: entering vehicles and leaving vehicles. After receiving this information, the planning module locates the path of the entering vehicle in the potential field map output in step 2 and pre-sets potential fields on both sides of the entering vehicle path. At this time, the potential fields generated by different vehicles are linearly superimposed, and a path search is performed from the entry point to the operation point to generate multiple candidate paths. From the output candidate paths, the optimal path is selected as the operation path through the uniform evaluation function.

[0076] In one embodiment, the method of presetting the potential field in step 31 includes:

[0077] When d car <d default When the potential field is g car [1-(d car / d default )], when d car ≥d default When , the potential field is 0.

[0078] Where g car The potential field value generated by the set vehicle path satisfies g car ≥g default , which can ensure that the impact of the vehicle potential field is greater than the impact of the initialization potential field, d car is the lateral offset of the projection of the point to the vehicle path, d default The influence distance is set, usually two vehicle widths.

[0079] For example, when vehicle B plans its route, vehicle A generates Figure 5 To prevent incoming vehicles from encroaching on the path of outgoing vehicles, additional potential fields need to be applied in the concentrated area of ​​exit routes when planning the operation path, such as Figure 5 Midpoint P c The farthest point P in the upper part of the region up The additional potential field between them. Its size is set to k*g car , where k is the number of operating points (P1, P2, P3) to ensure that the incoming vehicle will not plan a path that invades the exit area due to avoiding other vehicles.

[0080] Those skilled in the art may also adopt other expressions for attenuation with distance, which may achieve the same effect as the potential field preset in the above embodiment.

[0081] Step 32, planning the exit path: Clear the potential field generated by the incoming vehicle during the operation path planning. At the same time, based on the status information of other vehicles provided by the background, locate the path of the outgoing vehicle in the potential field map output in step 2, and pre-set potential fields on both sides of its path. Search for a path from the operation point to the exit point to generate the exit path.

[0082] In one embodiment, the method of presetting the potential field in step 32 includes:

[0083] When planning the exit path, an additional potential field is applied to the concentrated area of ​​the entry route to prevent the entry vehicle from occupying the exit position or the exit vehicle from occupying the entrance position, which is used to guide the planning module to search the path to one side. Figure 6 Click P c and the farthest point P in the lower and upper part of the region down The additional potential field between .

[0084] However, the above methods cannot solve the deadlock problem. The reversing section before entering the work point is a high-frequency section for deadlock. This is because when a vehicle is in reverse, if the path is occupied by a vehicle in forward state, it is bound to cause the vehicle to go in the opposite direction. This situation cannot be avoided by speed planning. Therefore, whether planning the work path or the exit path, the planning module will set a higher potential field for the reversing section of other vehicles in the area, with a size of g obs , to ensure that there is no spatial conflict in each reversing section, such as Figure 6 The potential field formed on the reversing road section near the operating points P1 and P3 is an artificial potential field, similar to the magnetic field repulsion in physics and the electric potential energy in electricity.

[0085] In one embodiment, the evaluation method of candidate paths and uniform operation planning: If the operation point in the operation area is idle for a long time, it will lead to a waste of resources, and long-term use will lead to irregular expansion of the operation area. Therefore, the selection of appropriate operation points by the operation vehicle is the key to uniform operation. In order to ensure the balance of the number of visits to each operation point, the background scheduling system will record the number of visits to the operation point num, and the vehicle planning module will select the best operation point based on the number of visits to the operation point. Specifically, the vehicle planning module generates i paths with each operation point as the planning end point in turn under the potential field condition, such as Figure 7 In this embodiment, step 4 specifically includes:

[0086] Step 41: Sample m on the i-th candidate path with a spacing of Δs i path points, and the potential field values ​​of all path points are accumulated to obtain the total potential field value of the path. If the path planning fails, Set to positive infinity, and calculate the evaluation value of the i-th candidate path through the following evaluation function

[0087]

[0088] Where:

[0089] The additional potential field for the number of visits to the kth operating point is used to ensure that the number of visits to each operating point tends to be consistent, which is the main influencing indicator of uniform operation. It can be expressed as: It can also be expressed as a potential field value that is linearly superimposed with the number of visits, and can be further expanded on this basis to obtain other forms

[0090] The time-attenuated additional potential field for the kth operating point is used to adjust the working interval of the operating point and prevent vehicles from waiting near the operating point and idling for a long time. It can be expressed as: certainly, It can also be expressed in other forms as long as it can be guaranteed to decay over time, where t is the current moment and t operate is the time when the last vehicle entered the operation point, and ΔT is the default time decay interval, which is usually set to 60s.

[0091] Here, Δs is typically 0.1m. Furthermore, path planning failure can be determined by, for example, the search algorithm iteration time. If the search algorithm iteration time is too long, for example, exceeding 2s, planning is considered a failure. Besides using the aforementioned expression, the evaluation function can also be one that considers the comprehensive potential field of waypoints, the cumulative number of workpoint visits, and the decay of the visit time. Other existing evaluation functions are also possible.

[0092] Step 42, select The smallest path is taken as the optimal path.

[0093] An embodiment of the present invention further provides a multi-vehicle uniform loading and unloading path planning system for a mine dump, which includes:

[0094] The vehicle information acquisition module is used to obtain the status information, trajectory information and visit count of other vehicles in the operation area from the background after the vehicle enters the planned starting point;

[0095] The planning module is used to initialize the potential field map based on the map data and adjust the potential field of the potential field map according to the paths of other vehicles in the work area;

[0096] A path planning module, which is used to generate an operation path and an exit path through a planning method of a variable potential field;

[0097] The job completion judgment module is used to continue the potential field adjustment by the planning module after the job is completed;

[0098] Among them, the path planning module includes:

[0099] The operation path planning unit is used to classify vehicles in the operation area into two categories: entering vehicles and leaving vehicles. The paths of the entering vehicles are located in the potential field map output in step 2. Potential fields are pre-set on both sides of the paths of the entering vehicles. The potential fields generated by different vehicles are linearly superimposed. A path search is performed from the entry point to the operation point to generate multiple candidate paths. From the output candidate paths, the optimal path is selected as the operation path through the homogenization evaluation function.

[0100] The exit path planning unit is used to clear the potential field generated by the incoming vehicle during the operation path planning. At the same time, based on the status information of other vehicles provided by the background, the path of the outgoing vehicle is located in the potential field map output by the planning module, and potential fields are pre-set on both sides of its path. The path from the operation point to the exit point is searched to generate the exit path.

[0101] In one embodiment, the method for presetting the potential field in the operation path planning unit includes:

[0102] When d car <d default When the potential field is g car [1-(d car / d default )], when d car ≥d default When , the potential field size is 0;

[0103] Where g car The potential field value generated by the set vehicle path satisfies g car ≥g default , which can ensure that the impact of the vehicle potential field is greater than the impact of the initialization potential field, d car is the lateral offset of the projection of the point to the vehicle path, d default The influence distance of the setting.

[0104] In one embodiment, the method for presetting the potential field in the exit path planning unit includes:

[0105] When planning the exit route, an additional potential field is applied in the entry route concentration area to prevent incoming vehicles from occupying the exit position or outgoing vehicles from occupying the entry position.

[0106] In one embodiment, the planning module specifically includes:

[0107] First find the farthest point P from the boundary in the working area c , P c The distance from the boundary is set to d k ; Then find the distance d from the boundary in the working area k -β point set: {P 11 ,P 12 ,....,P 1m}, then find the distance d from the boundary of the working area in the working area. k -2β point set: {P 21 ,P 22 ,....,P 2m}, repeat this process to find the distance d from the boundary in the working area k -nβ point set {P n1 ,P n2 ,....,P nm} point set; where β is the potential field attenuation distance and can be optimized according to the area of ​​the working area, m is the number of discrete points on the potential field boundary and is determined by the point set spacing, n is the number of potential field layers and is determined by the area of ​​the working area, and the regional center potential field of the annular artificial potential field is g default , based on the distance point P c The distance decreases in turn, and the potential field value decreases to 0 in the blank position around the center of the area. The vehicle needs to maintain a collision avoidance distance with the boundary of the circular artificial potential field. The potential field near the boundary is set to g obs .

[0108] In one embodiment, the path planning module specifically includes:

[0109] Sample m on the i-th candidate path with a spacing of Δs i path points, and the potential field values ​​of all path points are accumulated to obtain the total potential field value of the path. If the path planning fails, Set to positive infinity, and calculate the evaluation value of the i-th candidate path through the following evaluation function choose The smallest path is taken as the optimal path;

[0110]

[0111] Where, The additional potential field for the number of visits to the kth operating point is used to ensure that the number of visits to each operating point tends to be consistent. It is the main influencing indicator of uniform operation. The time-attenuated additional potential field for the kth operating point is used to adjust the working interval of the operating point and prevent vehicles from waiting near the operating point and idling for a long time.

[0112] Currently, there are few methods for multi-vehicle path planning in work areas. This paper designs a comprehensive multi-vehicle path planning method for work areas, including spatial avoidance of multi-vehicle paths, rational selection of work points, and uniform vehicle operation. Compared with numerical optimization methods, this method has faster computation speed and higher stability and engineering applicability than complete graph search methods. Overall, it is an algorithm with high real-time performance, security, and practical applicability.

[0113] Using other different rules to generate a potential field to guide the vehicle to plan a path in a specific direction, or changing the path evaluation function can achieve similar planning effects, but both belong to the main idea of ​​the present invention.

[0114] Finally, it should be noted that the above embodiments are intended only to illustrate the technical solutions of the present invention and are not intended to limit them. Those skilled in the art will appreciate that the technical solutions described in the aforementioned embodiments may be modified, or some of the technical features thereof may be replaced with equivalents; such modifications or replacements do not deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for planning uniform loading and unloading paths for multiple vehicles in a mine dump, characterized in that: include: Step 1: After entering the planned starting point, the vehicle obtains the status information, trajectory information and visit count of other vehicles in the operation area from the background; Step 2: Initialize the potential field map based on the map data and adjust the potential field of the potential field map according to the paths of other vehicles in the work area; Step 3: Generate the operation path and exit path through the planning method of the variable potential field; Step 4: After the job is completed, return to step 2; Among them, step 3 specifically includes: If it is a planned operation path, go to step 31, otherwise go to step 32; Step 31, planning the work path: vehicles in the work area are divided into two categories: entering vehicles and leaving vehicles. The paths of the entering vehicles are located in the potential field map output in step 2. Potential fields are pre-set on both sides of the paths of the entering vehicles. The potential fields generated by different vehicles are linearly superimposed. A path search is performed from the entry point to the work point to generate multiple candidate paths. From the output candidate paths, the optimal path is selected using a uniform evaluation function as the work path. Step 32, planning the exit path: Clear the potential field generated by the incoming vehicle during the operation path planning. At the same time, based on the status information of other vehicles provided by the background, locate the path of the outgoing vehicle in the potential field map output in step 2, and pre-set potential fields on both sides of its path. Search for a path from the operation point to the exit point to generate the exit path.

2. The method for planning paths for uniform loading and unloading of multiple vehicles at a mine dump as claimed in claim 1, characterized in that: The method for presetting the potential field in step 31 includes: When d car <d default When the potential field is g car [1-(d car / d default )], when d car ≥d default When , the potential field size is 0; Where g car The potential field value generated by the set vehicle path satisfies g car ≥g default , which can ensure that the impact of the vehicle potential field is greater than the impact of the initialization potential field, g default is the regional central potential field of the annular artificial potential field, d car is the lateral offset of the projection of the point to the vehicle path, d default The influence distance of the setting.

3. The method for planning paths for uniform loading and unloading of multiple vehicles at a mine dump as claimed in claim 1, wherein: The method for presetting the potential field in step 32 includes: When planning the exit route, an additional potential field is applied in the entry route concentration area to prevent incoming vehicles from occupying the exit position or outgoing vehicles from occupying the entry position.

4. The method for planning paths for uniform loading and unloading of multiple vehicles at a mine dump according to any one of claims 1 to 3, characterized in that: Step 2 specifically includes: First find the farthest point P from the boundary in the working area c , P c The distance from the boundary is set to d k ; Then find the distance d from the boundary in the working area k -β point set: {P 11 ,P 12 ,....,P 1m }, then find the distance d from the boundary of the working area in the working area. k -2β point set: {P 21 ,P 22 ,....,P 2m }, repeat this process to find the distance d from the boundary in the working area k -nβ point set {P n1 ,P n2 ,....,P nm } point set; where β is the potential field attenuation distance and can be optimized according to the area of ​​the working area, m is the number of discrete points on the potential field boundary and is determined by the point set spacing, n is the number of potential field layers and is determined by the area of ​​the working area, and the regional center potential field of the annular artificial potential field is g default , based on the distance P c The distance decreases in turn, and the potential field value decreases to 0 in the blank position around the center of the area. The vehicle needs to maintain a collision avoidance distance with the boundary of the circular artificial potential field. The potential field near the boundary is set to g obs .

5. The method for planning paths for uniform loading and unloading of multiple vehicles at a mine dump as claimed in claim 4, characterized in that: Step 4 specifically includes: Step 41: Sample m on the i-th candidate path with a spacing of Δs i path points, and the potential field values ​​of all path points are accumulated to obtain the total potential field value of the path. If the path planning fails, Set to positive infinity, and calculate the evaluation value of the i-th candidate path through the following evaluation function Where, The additional potential field for the number of visits to the kth operating point is used to ensure that the number of visits to each operating point tends to be consistent. It is the main influencing indicator of uniform operation. The time-attenuated additional potential field for the kth operating point is used to adjust the working interval of the operating point to prevent vehicles from waiting near the operating point and idling for a long time; Step 42, select The smallest path is taken as the optimal path.

6. A multi-vehicle uniform loading and unloading path planning system for a mine dump, characterized in that: include: The vehicle information acquisition module is used to obtain the status information, trajectory information and visit count of other vehicles in the operation area from the background after the vehicle enters the planned starting point; The planning module is used to initialize the potential field map based on the map data and adjust the potential field of the potential field map according to the paths of other vehicles in the work area; A path planning module, which is used to generate an operation path and an exit path through a planning method of a variable potential field; The job completion judgment module is used to continue the potential field adjustment by the planning module after the job is completed; Among them, the path planning module includes: The operation path planning unit is used to classify vehicles in the operation area into two categories: entering vehicles and leaving vehicles. The paths of the entering vehicles are located in the potential field map output in step 2. Potential fields are pre-set on both sides of the paths of the entering vehicles. The potential fields generated by different vehicles are linearly superimposed. A path search is performed from the entry point to the operation point to generate multiple candidate paths. From the output candidate paths, the optimal path is selected as the operation path through the homogenization evaluation function. The exit path planning unit is used to clear the potential field generated by the incoming vehicle during the operation path planning. At the same time, based on the status information of other vehicles provided by the background, the path of the outgoing vehicle is located in the potential field map output by the planning module, and potential fields are pre-set on both sides of its path. The path is searched from the operation point to the exit point to generate the exit path.

7. The multi-vehicle uniform loading and unloading path planning system for a mine dump as claimed in claim 6, characterized in that: Methods for presetting the potential field in the operation path planning unit include: When d car <d default When the potential field is g car [1-(d car / d default )], when d car ≥d default When , the potential field size is 0; Where g car The potential field value generated by the set vehicle path satisfies g car ≥g default , which can ensure that the impact of the vehicle potential field is greater than the impact of the initialization potential field, g default is the regional central potential field of the annular artificial potential field, d car is the lateral offset of the projection of the point to the vehicle path, d default The influence distance of the setting.

8. The multi-vehicle uniform loading and unloading path planning system for a mine dump as claimed in claim 6, characterized in that: Methods for presetting the potential field in the exit path planning unit include: When planning the exit route, an additional potential field is applied in the entry route concentration area to prevent incoming vehicles from occupying the exit position or outgoing vehicles from occupying the entry position.

9. The multi-vehicle uniform loading and unloading path planning system for a mine dump as claimed in any one of claims 6 to 8, characterized in that: The planning module specifically includes: First find the farthest point P from the boundary in the working area c , P c The distance from the boundary is set to d k ; Then find the distance d from the boundary in the working area k -β point set: {P 11 ,P 12 ,....,P 1m }, then find the distance d from the boundary of the working area in the working area. k -2β point set: {P 21 ,P 22 ,....,P 2m }, repeat this process to find the distance d from the boundary in the working area k -nβ point set {P n1 ,P n2 ,....,P nm } point set; where β is the potential field attenuation distance and can be optimized according to the area of ​​the working area, m is the number of discrete points on the potential field boundary and is determined by the point set spacing, n is the number of potential field layers and is determined by the area of ​​the working area, and the regional center potential field of the annular artificial potential field is g default , based on the distance P c The distance decreases in turn, and the potential field value decreases to 0 in the blank position around the center of the area. The vehicle needs to maintain a collision avoidance distance with the boundary of the circular artificial potential field. The potential field near the boundary is set to g obs .

10. The multi-vehicle uniform loading and unloading path planning system for a mine dump as claimed in claim 9, characterized in that: The path planning module specifically includes: Sample m on the i-th candidate path with a spacing of Δs i path points, and the potential field values ​​of all path points are accumulated to obtain the total potential field value of the path. If the path planning fails, Set to positive infinity, and calculate the evaluation value of the i-th candidate path through the following evaluation function choose The smallest path is taken as the optimal path; Where, The additional potential field for the number of visits to the kth operating point is used to ensure that the number of visits to each operating point tends to be consistent. It is the main influencing indicator of uniform operation. The time-attenuated additional potential field for the kth operating point is used to adjust the working interval of the operating point and prevent vehicles from waiting near the operating point and idling for a long time.

Images