A collaborative decision-making planning method for unmanned transport vehicles and shovels in well working conditions

Abstract

The present invention provides a method for collaborative decision-making and planning of unmanned transport vehicles and shovels in underground mining areas. The method realizes automatic planning of loading and unloading tasks of mining trucks and shovel loader operations through the interaction of a collaborative management and control platform with intelligent mining trucks and intelligent shovel loader. In the process of mining truck and shovel loader tasks, global path planning can be realized. When encountering obstacles, local paths can be automatically planned to avoid obstacles. This method can fully realize unmanned operation, and the collaboration of vehicles and shovels can improve the efficiency and safety of underground mining. Both mining trucks and shovel loader vehicles can automatically plan paths according to actual conditions.

Description

Technical Field

[0001] The present invention mainly relates to the technical field related to unmanned transportation in mines, and specifically to a collaborative decision-making and planning method for unmanned transportation vehicles and shovels in well-condition mining areas. Background Art

[0002] my country boasts a vast territory and abundant mineral resources, with proven reserves of 159 minerals and over 100,000 mines. Mining and transportation operations are a major concern due to environmental, safety, and efficiency issues. To address production pain points such as safety accidents in mining areas, difficulties recruiting drivers, and high management and operating costs, the development of smart mines and the rapid development of autonomous driving technology have led to the emergence of autonomous mining. Typically, mining operations include drilling, blasting, mining and loading, transportation, and dumping. The dumping process involves the safe positioning of autonomous vehicles within the mine, with the vehicles following a planned route to the unloading area to unload the ore. Autonomous mining vehicles primarily refer to autonomous mining trucks and loaders. As sub-modules of smart mines, they address transportation issues within the mine and are directly impactful to mining efficiency.

[0003] There are two types of mines: open-pit mines and underground mines. Open-pit mines are those within 200 meters of the surface. Digging typically begins at the surface, removing the soil before drilling downward in circles. However, underground mines are different. Over 85% of mines in my country, both coal and non-coal, are located underground. Underground light is dim, making traditional camera-based solutions ineffective. There's also a lack of signal signal underground, making traditional RTK ineffective. Implementing unmanned operations in these conditions is challenging, especially given the need for the truck and shovel to collaborate effectively and achieve planning and decision-making capabilities.

[0004] Therefore, it is particularly important to overcome the shortcomings of the existing technology and provide a method that can achieve efficient collaborative work and automatic path planning of intelligent mining trucks and scrapers in a well working environment. Summary of the Invention

[0005] In order to address the shortcomings of current technology, the present invention combines existing technology and, based on practical applications, provides a collaborative decision-making and planning method for unmanned transport vehicles and shovels in underground mining areas. The method has the advantages of fully realizing unmanned operation, improving the efficiency and safety of underground mining through vehicle-shovel collaboration, and automatically planning the paths of both mining trucks and shovel trucks according to actual conditions.

[0006] The technical solutions of the present invention are as follows:

[0007] A collaborative decision-making and planning method for unmanned transport trucks and shovels in well-operated mining areas is proposed. This method realizes the automatic planning of mining truck loading and unloading tasks and shovel operation through the interaction between the collaborative management and control platform, intelligent mining trucks and intelligent shovels.

[0008] The method for planning mining truck loading tasks is as follows:

[0009] S11. After the mining truck is started, it automatically registers the vehicle information with the platform, uploads the vehicle's location, task status, and chassis data, and requests the platform to issue tasks;

[0010] S12. After receiving the vehicle information, the platform automatically issues the corresponding task based on the data fed back by the vehicle. When the vehicle is unloaded, the platform will issue a loading task.

[0011] S13. After receiving the task, the mining truck analyzes the task and plans a global route to the loading area;

[0012] S14: After the mining truck arrives at the loading area, it waits for the platform to designate a loading point. The vehicle will then plan a route to the loading point, drive to the loading point, and wait for loading to be completed.

[0013] The scraper planning method is as follows:

[0014] S21. After the scraper is started, it automatically registers with the platform and uploads the vehicle location, speed, and status;

[0015] S22. After receiving the registration information of the scraper, the platform will display the scraper information and issue a loading task to the scraper after receiving the loading status of the mining truck;

[0016] S23: After receiving the loading task, the scraper parses the task and plans a global path to the vehicle-shovel collaboration area;

[0017] S24: After the scraper arrives at the vehicle-shovel collaboration area, it waits for the platform to designate a collaborative loading point. The scraper then plans a route to the collaborative loading point and drives to the collaborative loading point.

[0018] S25: When the scraper arrives at the collaborative loading point, it will automatically lift up. After reaching the top, there will be a delay. Then the ore in the bucket will be unloaded into the bucket of the mining truck. Then the bucket will be dropped. At this point, the scraper unloading is completed and it will leave the truck-shovel collaborative area.

[0019] The mining truck unloading task planning method is as follows:

[0020] S31. The platform judges the task status uploaded by the mining card and sends the unloading task to the mining card after it determines that the loading is completed.

[0021] S32: After receiving the unloading task, the mining truck will plan a global route to the unloading waiting area and drive to the unloading waiting area;

[0022] S33. The platform determines the designated unloading point when the mining truck reaches the unloading waiting area based on the real-time location information uploaded by the mining truck.

[0023] S34: After receiving the unloading point, the mining truck plans a route to the unloading point and drives to the unloading point;

[0024] S35: The mining truck drives to the unloading point, automatically lifts the bucket to unload the ore, and automatically drops the bucket after unloading is completed.

[0025] Furthermore, smart mining trucks and smart shovel loaders are positioned using SLAM positioning algorithms, and their locations are uploaded to the collaborative management and control platform in real time.

[0026] Furthermore, the tasks issued by the platform to mining trucks and shovel loaders include task road section information. When planning the path, the path planning modules on the mining trucks and shovel loaders use the task road section in the task as the global path for tracking. During driving, when the vehicle-side perception module senses that there are obstacles within the set range ahead, it will perform local path planning by scattering points horizontally and vertically.

[0027] Furthermore, the specific method of local path planning is as follows:

[0028] (a) Based on the positions of the four corner points of the obstacle given by the perception module, the vector cross product formula is used to calculate the cross product of the four corner points and the current trajectory point. The distance formula is used to calculate the distance between the four corner points of the obstacle and the reference trajectory. If the cross product results calculated by the four corner points have different signs, the obstacle is judged to be on the reference trajectory. If the obstacle is on one side of the trajectory, if the minimum distance from the four corner points to the reference trajectory is less than the set threshold, it is considered to be on the reference trajectory. Otherwise, it is considered not on the reference trajectory.

[0029] (b) When the obstacle is on the reference trajectory, the vertical and horizontal points are scattered. The vertical range is 0-40 meters, the spacing is 5 meters, and the horizontal range is the left and right road boundaries, the spacing is 0.5 meters.

[0030] (c) After the points are scattered, a fifth-order polynomial is used for fitting to solve multiple driving trajectories;

[0031] (d) After fitting the driving trajectory, the cost function is used to screen the optimal trajectory. The cost function takes into account factors such as lateral deviation, trajectory curvature, and trajectory impact when designing. When the calculated cost function is minimum, the trajectory is considered to be the optimal local path.

[0032] Furthermore, during local path planning, the calculation formula for determining whether an obstacle is on the trajectory is as follows:

[0033] Front left

[0034] fl_x = obstacle_info.at(i).point_front_left.x - path->poses[j].pose.position.x

[0035] fl_y = obstacle_info.at(i).point_front_left.y - path->poses[j].pose.position.y

[0036]

[0037] fl_cross = obstacle_info.at(i).point_front_left.x * P_y - obstacle_info.at(i).point_front_left.y * P_x front right

[0038] fr_x = obstacle_info.at(i).point_front_right.x - path->poses[j].pose.position.x

[0039] fr_y = obstacle_info.at(i).point_front_right.y - path->poses[j].pose.position.y

[0040]

[0041] fr_cross = obstacle_info.at(i).point_front_right.x * P_y - obstacle_info.at(i).point_front_right.y * P_x rear left

[0042] rl_x = obstacle_info.at(i).point_rear_left.x - path->poses[j].pose.position.x

[0043] rl_y = obstacle_info.at(i).point_rear_left.y - path->poses[j].pose.position.y

[0044]

[0045] rl_cross=obstacle_info.at(i).point_rear_left.x*P_y-obstacle_info.at(i).point_rear_left.y*P_x rear right

[0046] rr_x=obstacle_info.at(i).point_rear_right.x-path->poses[j].pose.position.x

[0047] rr_y=obstacle_info.at(i).point_rear_right.y-path->poses[j].pose.position.y

[0048]

[0049] rr_cross=obstacle_info.at(i).point_rear_right.x*P_y-obstacle_info.at(i).point_rear_right.y*P_x

[0050] In the above formula, fl_x, fl_y, dis_fl, and fl_cross represent the x-direction distance, y-direction distance, straight-line distance, and cross product result of the left front point of the i-th obstacle from the j-th trajectory point, respectively; fr_x, fr_y, dis_fr, and fr_cross represent the x-direction distance, y-direction distance, straight-line distance, and cross product result of the right front point of the i-th obstacle from the j-th trajectory point, respectively; rl_x, rl_y, dis_rl, and rl_cross represent the x-direction distance, y-direction distance, straight-line distance, and cross product result of the left rear point of the i-th obstacle from the j-th trajectory point, respectively. ; rr_x, rr_y, dis_rr, rr_cross respectively represent the x-direction distance, y-direction distance, straight-line distance, and cross product result of the left rear point of the i-th obstacle to the j-th trajectory point; obstacle_info.at(i).point_front_left.x and obstacle_info.at(i).point_front_left.y represent the x and y of the left front corner of the obstacle; obstacle_info.at(i).point_front_right.x,

[0051] obstacle_info.at(i).point_front_right.y represents the x and y values ​​of the right front corner of the obstacle; obstacle_info.at(i).point_rear_left.x and obstacle_info.at(i).point_rear_left.y represent the x and y values ​​of the left rear corner of the obstacle;

[0052] obstacle_info.at(i).point_rear_right.x and obstacle_info.at(i).point_rear_right.y represent the x and y coordinates of the right rear corner of the obstacle; path->poses[j].pose.position.x and path->poses[j].pose.position.y represent the x and y coordinates of the points on the path.

[0053] Furthermore, the cost function formula is as follows:

[0054]

[0055] In the formula, J represents the cost function, k1, k2, k3 represent the cost coefficients, represents the sum of squares of the curvature of the trajectory; ∑error_d i represents the lateral deviation of the trajectory and; represents the sum of squared jerk of the trajectory.

[0056] Furthermore, in step S14, the mining truck passes through the loading waiting area before entering the loading area. After entering the loading waiting area, the mining truck stops and applies to the platform to enter the loading area. The platform determines whether there are other vehicles in the loading area for loading tasks. If so, the mining truck is made to wait in the loading waiting area. If not, the platform sends a command to the mining truck to enter the loading area.

[0057] In step S24, the scraper passes through the collaborative waiting area before entering the vehicle-shovel collaborative area. After entering the collaborative waiting area, the scraper stops and applies to the platform to enter the vehicle-shovel collaborative area. The platform determines whether there are other vehicles performing loading tasks in the vehicle-shovel collaborative area. If so, the scraper is ordered to wait in the collaborative waiting area. If not, a command is issued to the scraper to enter the vehicle-shovel collaborative area.

[0058] Beneficial effects of the present invention:

[0059] 1. The planning method provided by the present invention can realize the automatic allocation of tasks and the automatic operation of mining trucks and scrapers through the interaction between the unmanned system management and control platform and the intelligent mining trucks and scrapers. Through the collaborative operation of mining trucks and scrapers, the efficiency and safety of unmanned mining in well conditions can be greatly improved.

[0060] 2. In the present invention, the mining truck and the scraper can realize automatic path planning and automatically avoid obstacles when encountering obstacles. It has a high degree of intelligence and is suitable for use in well working environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is a diagram of the vehicle-shovel collaborative architecture of the present invention.

[0062] Figure 2 Loading task flow chart for mining trucks.

[0063] Figure 3 Run the flow chart for the scraper.

[0064] Figure 4 Flowchart of the task of unloading mining cards. DETAILED DESCRIPTION

[0065] The present invention will be further described with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are intended only to illustrate the present invention and are not intended to limit the scope of the present invention. In addition, it should be understood that after reading the contents of the present invention, those skilled in the art may make various changes or modifications to the present invention, and these equivalent forms also fall within the scope defined by the present application.

[0066] This embodiment provides a collaborative decision-making and planning method for unmanned transport vehicles and shovels in a well-operated mining area.

[0067] refer to Figure 1 The above is a flowchart of the entire vehicle-shovel collaborative architecture. The unmanned collaborative management and control platform (hereinafter referred to as the platform) is primarily responsible for displaying vehicle status, vehicle mileage, task scheduling, and operation records. In this embodiment, the collaborative management and control platform interacts with smart mining trucks and smart scrapers to implement automatic loading and unloading planning for the mining trucks and operation planning for the scrapers.

[0068] refer to Figure 2 As shown, the workflow of the mining truck automatic loading task provided in this embodiment is as follows.

[0069] S11. After the mining truck is started, it automatically registers its vehicle information with the platform, uploads its location, mission status, and chassis data, and requests the platform to issue a mission. As a preferred solution, in this embodiment, both the mining truck and the scraper are positioned using a SLAM positioning algorithm.

[0070] S12. After receiving the vehicle information, the platform automatically issues the corresponding task based on the data fed back by the vehicle. When the vehicle is unloaded, the platform will issue a loading task.

[0071] S13. After receiving a task, the vehicle analyzes it. The task content includes the number of tasks, task sections, task type, task subtype, and stopping points. When there are no obstacles, the vehicle's path planning module uses the task sections as the global path for tracking. When the perception module detects an obstacle within 30 meters, it performs local trajectory planning using a lattice method. This lattice method involves scattering points at a certain distance in front of the vehicle, such as 5, 10, or 15 meters longitudinally, with the vehicle's centerline as the reference point. This is called the lattice algorithm.

[0072] In this embodiment, a local path planning method is provided. This method is also used in scraper operation and mining truck unloading tasks. The specific local path planning steps are as follows:

[0073] (a) Based on the perceived positions of the four corner points of the obstacle, use the vector cross product formula to calculate the cross product of the four corner points and the current trajectory point; use the distance formula to calculate the distance between each corner point of the obstacle and the reference trajectory. If the cross product results calculated from the four corner points have different signs, it proves that the obstacle is on the reference trajectory. Secondly, if the obstacle is on one side of the trajectory, the minimum distance from the four corner points to the reference line is determined. If it is less than the threshold, it is considered to be on the reference trajectory. Otherwise, it is not on the reference trajectory. The formula for determining whether the obstacle is on the trajectory is as follows:

[0074] Front left

[0075] fl_x=obstacle_info.at(i).point_front_left.x-path->pos es[j].pose.position.x

[0076] fl_y=obstacle_info.at(i).point_front_left.y-path->pos es[j].pose.position.y

[0077]

[0078] fl_cross=obstacle_info.at(i).point_front_left.x*P_y-obstacle_info.at(i).point_front_left.y*P_xfront right

[0079] fr_x = obstacle_info.at(i).point_front_right.x - path->poses[j].pose.position.x

[0080] fr_y = obstacle_info.at(i).point_front_right.y - path->poses[j].pose.position.y

[0081]

[0082] fr_cross = obstacle_info.at(i).point_front_right.x * P_y - obstacle_info.at(i).point_front_right.y * P_x rear left

[0083] rl_x = obstacle_info.at(i).point_rear_left.x - path->poses[j].pose.position.x

[0084] rl_y = obstacle_info.at(i).point_rear_left.y - path->poses[j].pose.position.y

[0085]

[0086] rl_cross = obstacle_info.at(i).point_rear_left.x * P_y - obstacle_info.at(i).point_rear_left.y * P_x rear right

[0087] rr_x = obstacle_info.at(i).point_rear_right.x - path->poses[j].pose.position.x

[0088] rr_y = obstacle_info.at(i).point_rear_right.y - path->poses[j].pose.position.y

[0089]

[0090] rr_cross=obstacle_info.at(i).point_rear_right.x*P_y-obstacle_info.at(i).point_rear_right.y*P_x

[0091] Where: fl_x, fl_y, dis_fl, and fl_cross represent the x-direction distance, y-direction distance, straight-line distance, and cross product result of the left front point of the i-th obstacle from the j-th trajectory point, respectively; fr_x, fr_y, dis_fr, and fr_cross represent the x-direction distance, y-direction distance, straight-line distance, and cross product result of the right front point of the i-th obstacle from the j-th trajectory point, respectively; rl_x, rl_y, dis_rl, and rl_cross represent the x-direction distance, y-direction distance, straight-line distance, and cross product result of the left rear point of the i-th obstacle from the j-th trajectory point, respectively; rr_x, rr_y, dis_rr, and rr_cross represent the x-direction distance, y-direction distance, straight-line distance, and cross product result of the left rear point of the i-th obstacle to the j-th trajectory point, respectively; obstacle_info.at(i).point_front_left.x and obstacle_info.at(i).point_front_left.y represent the x and y of the left front corner of the obstacle; obstacle_info.at(i).point_front_right.x and

[0092] obstacle_info.at(i).point_front_right.y represents the x and y values ​​of the right front corner of the obstacle; obstacle_info.at(i).point_rear_left.x and obstacle_info.at(i).point_rear_left.y represent the x and y values ​​of the left rear corner of the obstacle;

[0093] obstacle_info.at(i).point_rear_right.x and obstacle_info.at(i).point_rear_right.y represent the x and y coordinates of the right rear corner of the obstacle; path->poses[j].pose.position.x and path->poses[j].pose.position.y represent the x and y coordinates of the points on the path.

[0094] (b) When the obstacle is on the reference track, it is necessary to spread the points horizontally and vertically. Since the mining truck runs at a slow speed, the longitudinal spread range is set to 0-40 meters, with a spread interval of 5 meters; the horizontal spread range is [left road boundary, right road boundary], with a spread interval of 0.5 meters;

[0095] After the points are scattered, a 5th-order polynomial is used for fitting, and then multiple driving trajectories are solved. The formula is as follows

[0096] f(s)=a0+a1s+a2s 2 +a3s 3 +a4s 4 +a5s 5

[0097] Where a0, a1, a2, a3, a4, and a5 are fitting parameters, and f(s) is a function relative to the longitudinal deviation S.

[0098] (c) After fitting the driving trajectory, a cost function is used to select the optimal driving trajectory. The cost function takes into account factors such as lateral deviation, trajectory curvature, and trajectory impact when designing. When the calculated cost function is the smallest, the trajectory is considered to be the optimal local path. The cost function is expressed as follows:

[0099]

[0100] Where: J represents the cost function, k1, k2, k3 represent the cost coefficients, represents the sum of squares of the curvature of the trajectory; ∑error_d i represents the lateral deviation of the trajectory and; represents the sum of squared jerk of the trajectory.

[0101] S14. Before reaching the loading area, the mining truck will pass through the loading waiting area. After entering the loading waiting area, the mining truck will stop and apply to the cloud control platform to enter the loading area. If the cloud control platform receives an instruction to enter, the mining truck will drive to the loading area. If the cloud control platform receives an instruction to prohibit entry, the mining truck will stop in the loading waiting area and wait for the platform to enter.

[0102] When the mining truck receives the instruction to enter, it will drive to the loading area according to the planned path; after the platform determines that the vehicle has arrived at the loading area, it will send the loading point to the vehicle end, and then the vehicle end will plan the path to the loading point, and then drive to the loading point and wait for loading to be completed.

[0103] refer to Figure 3 The following is the operation process of the scraper.

[0104] S21. After the scraper is started, it will register with the platform and upload the vehicle location, speed, and status to the platform.

[0105] S22. After the platform receives the registration information of the scraper, it will display the scraper information on the unmanned collaborative management and control platform, and after receiving the loading status of the mining truck, it will issue a loading task to the scraper.

[0106] S23. After receiving the loading task, the scraper will perform global path planning between the current position and the loading waiting area. If it encounters obstacles in the map, it will also perform local path planning to avoid the obstacles.

[0107] S24. After the vehicle arrives at the loading waiting area, it will determine whether there are other vehicles queuing in the loading area. If there are shovel loaders, they will queue in the loading waiting area. If not, the vehicle will enter the vehicle-shovel collaboration area and wait for the platform to designate a collaborative loading point with the mining truck.

[0108] S25. After receiving the information that the scraper has arrived at the vehicle-shovel cooperation area, the platform sends the cooperation loading point to the scraper. After receiving the point, the scraper will perform path planning and then drive to the cooperation loading point.

[0109] S26. When the scraper arrives at the cooperative loading point, it will automatically lift up. After it is lifted to the top, there will be a delay. Then the ore in the bucket will be unloaded into the bucket of the mining truck, and then the bucket will be dropped. At this point, the scraper unloading is completed.

[0110] S27. After the scraper loader completes unloading, it drives out of the vehicle-shovel collaboration area and enters the shoveling area to wait for platform instructions.

[0111] refer to Figure 4 The following is the unloading task flow of the mining truck.

[0112] S31. When the mining card waits for the platform to issue an unloading task, the platform makes a judgment based on the task status uploaded by the mining card. After determining that the loading is completed, the platform will issue the unloading task.

[0113] S32. After receiving the unloading task, the mining truck will plan a global path to the unloading waiting area;

[0114] S33. After receiving the flag indicating successful path planning, the vehicle drives along the global path. When it is determined that there is an obstacle ahead, the vehicle starts local path planning and plans an obstacle avoidance path. The vehicle avoids the obstacle and then returns to the global path to continue driving.

[0115] S34. The platform will determine the location uploaded by the mining truck. When it reaches the unloading waiting area, it will specify the unloading point. If it does not reach the unloading waiting area, it will continue to drive.

[0116] S35. After receiving the unloading point, the mining truck will plan the route to the unloading point and then drive to the unloading point.

[0117] S36. When the truck reaches the unloading point, it automatically starts to lift the bucket. The angle sensor is used to detect in real time whether the bucket lifting action is completed. When the set threshold is reached, that is, when the data detected by the angle sensor reaches the set threshold, the bucket lifting will be delayed to wait for the ore to be unloaded.

[0118] S37. After the ore is unloaded, the bucket will drop, which is also detected by the angle sensor. When the data is less than the set threshold, the bucket drop is completed.

[0119] S38: After the mining truck bucket is dropped off, unloading is completed and the task is finished.

[0120] Through the decision-making planning method of the present invention, the planning of the loading and unloading tasks of the mining truck and the operation of the scraper is realized, so that the mining truck and the scraper can automatically plan the path and work together to complete the automatic loading and unloading tasks of the ore, with reliable operation and high loading efficiency.

Claims

1. A collaborative decision-making and planning method for unmanned transport vehicles and shovels in a well-condition mining area. This method realizes the automatic planning of loading and unloading tasks of mining trucks and the operation of shovels and shovels by interacting with a collaborative management and control platform, intelligent mining trucks and intelligent shovels and shovels, and is characterized in that: The method for planning the mining truck loading task is as follows: S11. After the mining truck is started, it automatically registers the vehicle information to the platform, uploads the vehicle location, task status, and chassis data, and requests the platform to issue tasks; S12. After receiving the vehicle information, the platform automatically issues the corresponding task based on the data fed back by the vehicle. When the vehicle is in an unloaded state, the platform will issue a loading task; S13. After receiving the task, the mining truck analyzes the task and plans a global path to the loading area; S14: After the mining truck arrives at the loading area, it waits for the platform to designate a loading point. The truck will plan a route to the loading point, then drive to the loading point and wait for loading to be completed. The scraper planning method is as follows: S21. After the scraper is started, it will automatically register with the platform and upload the vehicle location, speed, and status; S22. After receiving the registration information of the scraper, the platform will display the scraper information and issue a loading task to the scraper after receiving the loading status of the mining truck; S23, after receiving the loading task, the scraper parses the task and plans a global path to the vehicle-shovel cooperation area; S24, after the scraper arrives at the vehicle-shovel cooperation area, it waits for the platform to designate a cooperation loading point. The vehicle side will plan a path to the cooperation loading point and then drive to the cooperation loading point; S25, when the scraper arrives at the cooperative loading point, it will automatically lift up, and after lifting to the top, it will delay for a while, and then unload the ore in the bucket into the bucket of the mining truck, and then perform the bucket drop operation. At this point, the scraper unloading is completed, and after completion, it will drive out of the truck-shovel cooperative area; The method for planning the mining truck unloading task is as follows: S31. The platform makes a judgment based on the task status uploaded by the mining card. When it is judged that the loading is completed, it will send the unloading task to the mining card; S32: After receiving the unloading task, the mining truck will plan a global path to the unloading waiting area and drive to the unloading waiting area; S33, the platform determines the designated unloading point when the mining truck drives to the unloading waiting area based on the location information uploaded in real time by the mining truck; S34, after receiving the unloading point, the mining truck plans a route to the unloading point and drives to the unloading point; S35: The mining truck drives to the unloading point, automatically lifts the bucket to unload the ore, and automatically drops the bucket after unloading; The tasks sent by the platform to mining trucks and shovel trucks include task section information. When planning the path, the path planning module on the mining truck and shovel truck side takes the task section in the task as the global path for tracking. During the driving process, when the vehicle-side perception module senses that there are obstacles within the set range ahead, it will perform local path planning by scattering points horizontally and vertically. The specific method of local path planning is as follows: (a) Based on the positions of the four corner points of the obstacle given by the perception module, the vector cross product formula is used to calculate the cross product results of the four corner points and the current trajectory point, and the distance formula is used to calculate the distance from the four corner points of the obstacle to the reference trajectory. If the cross product results calculated by the four corner points have different signs, the obstacle is judged to be on the reference trajectory. If the obstacle is on one side of the trajectory, if the minimum distance from the four corner points to the reference trajectory is less than the set threshold, it is considered to be on the reference trajectory. Otherwise, it is considered not to be on the reference trajectory. (b) When the obstacle is on the reference track, the points are scattered horizontally and vertically. The longitudinal scattering range is 0-40 meters, the scattering interval is 5 meters, and the lateral scattering range is the left road boundary and the right road boundary, and the scattering interval is 0.5 meters. (c) After the points are scattered, a fifth-order polynomial is used for fitting to solve multiple driving trajectories; (d) After fitting the driving trajectory, the cost function is used to screen the optimal trajectory. The cost function takes into account the factors of lateral deviation, trajectory curvature, and trajectory impact when designing. When the calculated cost function is the smallest, the trajectory is considered to be the optimal local path; When planning a local path, the calculation formula for determining whether an obstacle is on the trajectory is as follows: Front left fl_x=obstacle_info.at(i).point_front_left.x-path->pos es[j].pose.position.x fl_y=obstacle_info.at(i).point_front_left.y-path->pos es[j].pose.position.y fl_cross=obstacle_info.at(i).point_front_left.x*P_y-obstacle_info.at(i).point_front_left.y*P_x Front right fr_x=obstacle_info.at(i).point_front_right.x-path->poses[j].pose.position.x fr_y=obstacle_info.at(i).point_front_right.y-path->poses[j].pose.position.y fr_cross=obstacle_info.at(i).point_front_right.x*P_y-obstacle_info.at(i).point_front_right.y*P_x Back left rl_x=obstacle_info.at(i).point_rear_left.x-path->poses[j].pose.position.x rl_y=obstacle_info.at(i).point_rear_left.y-path->poses[j].pose.position.y rl_cross=obstacle_info.at(i).point_rear_left.x*P_y-obstacle_info.at(i).point_rear_left.y*P_x rear right rr_x=obstacle_info.at(i).point_rear_right.x-path->poses[j].pose.position.x rr_y=obstacle_info.at(i).point_rear_right.y-path->poses[j].pose.position.y rr_cross=obstacle_info.at(i).point_rear_right.x*P_y-obstacle_info.at(i).point_rear_right.y*P_x In the above formula, fl_x, fl_y, dis_fl, and fl_cross represent the x-direction distance, y-direction distance, straight-line distance, and cross product result of the left front point of the ith obstacle from the jth trajectory point, respectively; fr_x, fr_y, dis_fr, and fr_cross represent the x-direction distance, y-direction distance, straight-line distance, and cross product result of the right front point of the ith obstacle from the jth trajectory point, respectively; rl_x, rl_y, dis_rl, and rl_cross represent the x-direction distance, y-direction distance, straight-line distance, and cross product result of the left rear point of the ith obstacle from the jth trajectory point, respectively. ; rr_x, rr_y, dis_rr, rr_cross represent the x-direction distance, y-direction distance, straight-line distance, and cross product result of the left rear point of the ith obstacle from the jth trajectory point, respectively; obstacle_info.at(i).point_front_left.x, obstacle_info.at(i).point_front_left.y represent the x and y of the left front corner of the obstacle; obstacle_info.at(i).point_front_right.x, obstacle_info.at(i).point_front_right.y represents the x and y values ​​of the right front corner of the obstacle; obstacle_info.at(i).point_rear_left.x and obstacle_info.at(i).point_rear_left.y represent the x and y values ​​of the left rear corner of the obstacle; obstacle_info.at(i).point_rear_right.x and obstacle_info.at(i).point_rear_right.y represent the x and y coordinates of the right rear corner of the obstacle; path->poses[j].pose.position.x and path->poses[j].pose.position.y represent the x and y coordinates of the points on the path.

2. The collaborative decision-making planning method for unmanned transport vehicles and shovels in a well working area according to claim 1 is characterized in that: Smart mining trucks and smart shovel loaders are positioned using SLAM positioning algorithms, and their locations are uploaded to the collaborative management and control platform in real time.

3. The collaborative decision-making planning method for unmanned transport vehicles and shovels in a well working area according to claim 1 is characterized in that: The cost function formula is as follows: In the formula, J represents the cost function, k1, k2, k3 represent the cost coefficients, represents the sum of the squares of the curvature of the trajectory; ∑error_d i represents the lateral deviation of the trajectory and; represents the sum of squared trajectory accelerations.

4. The collaborative decision-making planning method for unmanned transport vehicles and shovels in a well working area according to claim 1 is characterized in that: In step S14, the mining truck passes through the loading waiting area before entering the loading area. After entering the loading waiting area, the mining truck stops and applies to the platform to enter the loading area. The platform determines whether there are other vehicles in the loading area for loading tasks. If so, the mining truck is made to wait in the loading waiting area. If not, a command is issued to the mining truck to enter the loading area. In step S24, the shovel loader passes through the cooperation waiting area before entering the vehicle-shovel cooperation area. After entering the cooperation waiting area, the shovel loader stops and applies to the platform to enter the vehicle-shovel cooperation area. The platform determines whether there are other vehicles in the vehicle-shovel cooperation area to perform loading tasks. If so, the shovel loader is made to wait in the cooperation waiting area. If not, a command is issued to the shovel loader to enter the vehicle-shovel cooperation area.

Images