Self-driving mine car shovel position planning method

Abstract

The invention relates to an automatic driving mine car shovel position planning method which comprises the following steps: determining a theoretical loading area around a loading point, and removing an obstacle area from the theoretical loading area to obtain a loading area; generating a plurality of parking spaces in the loading area according to a set rule to form a candidate parking space set; generating a candidate reverse parking track set; and evaluating each reversing parking track in the candidate reversing parking track set according to a preset evaluation index, and selecting the reversing parking track of which the evaluation result meets a preset requirement as a target path. According to the shovel position planning method for the self-driving mine car, the most suitable coordinate and course angle of the unmanned vehicle parking on the excavator can be solved only based on the grid map and the coordinates of the excavator, and it is guaranteed that the unmanned mine car selects the most suitable loading position and the shortest backing-up and entering path; the method ensures the success rate of path planning of backing in, and aims to solve the problems of efficiency and reliability of a mine unmanned mine car loading scene.

Description

Technical Field

[0001] The present invention relates to the technical field of automatic driving path planning, and in particular to a method for planning a shovel position of an automatic driving mine vehicle. Background Art

[0002] Regarding the collaborative operation of excavators and mine trucks in open-pit mining, when a manned mine truck is operating, the driver observes the position of the excavator and the surrounding environment. When an unmanned mine truck cooperates with a manned excavator, the excavator needs to be manually operated to send the current excavator position coordinates and heading angle to the unmanned mine truck, so that the unmanned mine truck can calculate the path based on this information. However, existing technologies require that the excavator must stop rotating when sending the heading angle and wait for the combined inertial navigation to provide the current heading angle, which reduces the operating efficiency of the excavator and causes the overall production efficiency of the unmanned mine truck to be unable to reach the efficiency of manual driving.

[0003] When working with a manned excavator, the current unmanned mining vehicle is basically completely dependent on the manned excavator to provide the heading angle, which is the angle of the unmanned mining vehicle to park in front of the excavator. This given heading angle limits the vehicle's planned path and often leads to planning failures. Because the current commonly adopted solution is to obtain the coordinates and heading angle of the excavator, it is necessary to install a high-precision combined inertial navigation on the excavator, and the high-precision combined inertial navigation costs tens of thousands of yuan, which is costly. Summary of the invention

[0004] In view of the existing technical problems, the present invention provides a technical subject.

[0005] The technical solution of the present invention provides a method for planning the shovel position of an autonomous driving mine vehicle, comprising the following steps: S1. Determine the parking area for the mine car: obtain the location of the loading point, determine the theoretical loading area around the loading point, and remove the obstacle area from the theoretical loading area to obtain the loading area; S2, generating a candidate parking space set step: generating a plurality of parking spaces in the loading area according to set rules to form a candidate parking space set, wherein the parking spaces are represented by parking space coordinates and orientation angles; S3, step of generating a set of candidate reversing trajectories: generating a reversing trajectory for each parking space in the candidate parking space set to form a set of candidate reversing trajectories; S4, step of selecting a trajectory that meets the requirements as the target path: evaluating each reversing-into-garaging trajectory in the candidate reversing-into-garaging trajectory set with a preset evaluation index, and selecting the reversing-into-garaging trajectory whose evaluation result meets the preset requirements as the target path.

[0006] Preferably, in the step S1 of determining the mine car parking area, the theoretical loading area is determined based on the maximum operating radius, the minimum operating radius of the operating vehicle such as the excavator and the size of the mine car compartment.

[0007] Preferably, the maximum operating radius, the minimum operating radius and the size of the mine car compartment are determined based on the obtained loading equipment and the type of mine car.

[0008] Preferably, in the step S1 of determining the mine car stoppable area, a grid map of the surrounding environment of the loading point is obtained, the grid map and the theoretical loading area are superimposed, and then the obstacle area in the grid map is deleted from the theoretical loading area to determine the loading area.

[0009] Preferably, in the step S2 of generating a candidate parking space set, a number of parking spaces are determined according to a preset rule to obtain a parking space set; and parking spaces whose projection area of ​​the mining car exceeds the loading area are eliminated to obtain a candidate parking space set.

[0010] Preferably, in the step of eliminating parking spaces in which the projection area of ​​the mining car exceeds the loading area, a grid area is generated for each parking space based on the size of the mining car, the center point and the orientation angle, and it is determined whether the grid area is located in the loading area; only parking spaces in which the grid area completely belongs to the loading area are retained as the candidate parking space set.

[0011] Preferably, in the step of determining a number of parking spaces according to preset rules, a number of mine car center points are selected in the loading area, and then a number of vehicle orientation angles are generated at each mine car center point, and a number of parking spaces are obtained by combining the mine car center points and the orientation angles to form a parking space set.

[0012] Preferably, in the step of selecting a number of mine car center points, an optimal loading line is determined in the loading area, and a number of coordinate points are selected on the optimal loading line as the mine car center points.

[0013] Preferably, in the step of selecting a trajectory that meets the requirements as the target path in S4, the elements in the candidate reverse parking trajectory set are sorted from best to worst according to the evaluation result. The optimal parking trajectory that meets the target conditions is selected from the candidate reverse parking trajectory set as the target path.

[0014] Preferably, the step S4 of selecting a trajectory that meets the requirements as the target path determines the evaluation result based on the shortest path search comparison.

[0015] This method only needs to obtain the coordinates of the excavator and only needs a high-precision positioning module, with a cost as low as about a thousand yuan, which has obvious cost advantages. The automatic driving mine car shovel position planning method of the present invention does not rely on the heading angle provided by the excavator. It is only based on the grid map and the coordinates of the excavator on the grid map. The most suitable coordinates and heading angles for the unmanned vehicle to park on the excavator can be solved through the algorithm, and the unmanned mine car is guaranteed to select the most suitable loading position and the shortest reverse entry path, ensuring the success rate of the path planning for reverse entry, aiming to solve the efficiency and reliability of unmanned mine cars in loading scenarios in mines. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the overall process of the shovel position planning method of the automatic driving mine vehicle of the present invention; Figure 2 It is a schematic diagram of a specific process of the method for planning the shovel position of an autonomous driving mine vehicle of the present invention; Figure 3 A schematic diagram of a theoretical loading area of ​​the shovel position planning method for an autonomous mining vehicle of the present invention; Figure 4 A schematic diagram of a loading area of ​​the autonomous mining vehicle shovel position planning method of the present invention. DETAILED DESCRIPTION

[0017] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. In this specification, the size ratios in the drawings do not represent the actual size ratios, but are only used to reflect the relative position relationship and connection relationship between the components. Components with the same name or the same number represent similar or identical structures and are only for illustrative purposes.

[0018] The key to the shovel position planning method of the autonomous driving mine truck of the present application is to calculate the optimal loading position based on the position of the excavator and the surrounding environment, and ensure that the loading position is optimal. The core difference between it and the prior art is that it does not need to rely on the excavator driver to provide loading coordinates and vehicle orientation, and then generate the corresponding loading path based on the input. Its advantage is that the path planning can be completed in advance and shared by multiple subsequent loading tasks. Therefore, the purpose of reducing the amount of calculation and optimizing the operating efficiency is achieved.

[0019] The general process of the shovel position planning method of the autonomous driving mine car of this application is as follows Figure 1 , 2 As shown, it includes the following steps.

[0020] S1, determining the area where the mine car can stop. Obtain the location information of the loading point, and determine a theoretical loading area around the loading point according to the loading needs. Obtain a grid map around the loading point, and remove the obstacle area in the grid map from the theoretical loading area to obtain the loading area.

[0021] S2, step of generating a candidate parking space set: generating a plurality of parking spaces in the loading area according to set rules to form a candidate parking space set.

[0022] S3, generating a candidate reverse parking trajectory set step: generating a reverse parking trajectory for each parking space in the candidate parking space set to form a candidate reverse parking trajectory set.

[0023] S4, selecting a trajectory that meets the requirements as the target path step. Each reversing trajectory in the candidate reversing trajectory set is evaluated with a preset evaluation index, and the elements in the candidate reversing trajectory set are sorted from best to worst according to the evaluation result. The optimal reversing trajectory that meets the target condition is selected from the candidate reversing trajectory set as the target path.

[0024] In step S1, the mine car parking area is determined. The loading point can be considered as the coordinate point of an operating vehicle such as an excavator, so it can be determined by the RTK positioning information of the excavator, which can be determined by the high-precision positioning module installed on the excavator. Unless otherwise specified, all coordinate data and map data in this application are based on the CCS2000 Chinese Geodetic Coordinate System.

[0025] The theoretical loading area is determined based on the maximum operating radius, minimum operating radius and the size of the mine car compartment of the excavator and other operating vehicles. Basically, based on the above parameters in the actual project, it is only necessary to determine the maximum loading circle corresponding to the maximum loading radius and the minimum loading circle corresponding to the minimum loading radius. The circular area surrounded by the two is the theoretical loading area. According to the high-precision positioning module installed on the excavator, the real-time latitude and longitude information is obtained to determine its coordinates. The loading parameters can be determined by the received vehicle model information, and finally the theoretical loading area of ​​the excavator is obtained. Figure 3 The ring formed by the two concentric circles is the theoretical loading area of ​​the excavator bucket, and the middle circle is the optimal loading line, that is, when the center point of the unmanned carriage is on the optimal loading line, the loading range is optimal. However, there are multiple options for the orientation of the mine car at this time. Different orientations have different reversing routes. The parking space in this application not only refers to the parking space coordinates but also includes the corresponding orientation angle value of the mine car.

[0026] In the actual loading scenario, not all areas in the theoretical loading area are available. There are obstacle areas formed by the objects to be shoveled and other entities. These obstacle areas need to be removed to obtain an accurate loading area. The environmental features around the loading point can be extracted from the constructed grid map. Therefore, it is only necessary to overlay the grid map and the theoretical loading area, and then remove the obstacle areas in the grid map from the theoretical loading area to determine the loading area. Figure 4 When operating in a mine, the grid map can be directly generated based on the data collected by the LiDAR on the unmanned mining vehicle.

[0027] This process is preferably carried out in the vehicle coordinate system. Assume that the coordinates of the excavator based on the CGS2000 coordinate system are ( , ), the vehicle coordinate system is ( , , ), is the angle of the vehicle's front. At this time, the vehicle coordinate system is established with the vehicle coordinate as the origin and the vehicle orientation as the Y axis. At this time, the CGS2000 coordinates of the excavator are converted to ( , ), and the planning solution is performed in the vehicle coordinate system.

[0028] In the step S2 of generating a candidate parking space set, several candidate parking spaces are generated according to certain rules. During the generation process, it is necessary to ensure that the projection area of ​​the mine car is located in the loading area to avoid collisions between the mine car and obstacles. Specifically, several optional mine car center points can be determined in the loading area according to certain rules, and then several vehicle orientation angles are generated at each mine car center point. The mine car center points and orientation angles are combined to obtain several parking space sets, among which there are mine car parking spaces with different orientation angles at the same center point. The parking spaces obtained in this way do not take into account the volume of the mine car. In order to ensure that the mine car does not collide, it is necessary to generate a grid area for each parking space with the size of the mine car, the center point and the orientation angle, and determine whether the grid area is located in the loading area. Only the parking spaces whose grid areas are completely in the loading area are retained as the candidate parking space set.

[0029] The following is a specific example of determining a set of candidate parking spaces. On the loading centerline, a series of points are evenly spaced. ( ) as the minecart center. = { }, i=1, 2, 3 . As the center point, For the heading angle, a grid area of ​​the vehicle position and heading is generated. The point, whose coordinates are ( ), the orientation angle is , the vehicle is The coordinates and orientation angle of the point are ( ),according to , generates a grid area, the grid area is a two-dimensional array a, when the angle Add a Δ When the grid area changes, a new two-dimensional array is generated. . Determine whether the grid area is a parking area. If so, put it into the candidate parking space set. and A series of two-dimensional arrays a generated, a=[a1,a2,a3 ], calculate the intersection of each value of the two-dimensional array and the value of the grid map, and obtain a new array b with 0 and 1 as values. The sum of all elements in array b is B, where B represents the overlap between the carriage and the loading area. A threshold can be set according to the number of grids in the carriage. , when B and The closer it is, the more suitable the location is for parking and loading. According to the above judgment conditions, if the loading conditions are met, the coordinates and heading angle of the location are saved to generate a candidate parking space set C, where the candidate parking space set C={( )}.

[0030] S3 is a step of generating a set of candidate reverse parking trajectories. A reverse parking trajectory is generated for each parking space in the candidate parking space set to form a set of candidate reverse parking trajectories. For each candidate parking space in the candidate parking space set, a reverse parking trajectory is generated from a given starting point. In the actual application process, a hybrid A* algorithm is used to generate a reverse parking trajectory for each candidate parking space. The hybrid A* algorithm is a motion planning method that combines the traditional A* algorithm and the continuous state space. It aims to perform efficient path planning in robots such as vehicles with non-holonomic constraints (such as the turning radius of the vehicle). The Hybrid A* algorithm combines the ideas of the A* algorithm and the lattice planner algorithm, combines the path search of the grid map with the lattice map, selects different control quantities to preview a trajectory during the search process, and keeps only one feasible state in each grid. The technical solution of the present application does not exclude the use of other path planning methods to determine the reverse parking trajectory corresponding to each candidate parking space. The reverse parking trajectories of all candidate parking spaces constitute a set of candidate reverse parking trajectories.

[0031] S4, the step of selecting a trajectory that meets the requirements as the target path. Each reversing-into-garage trajectory in the candidate reversing-into-garage trajectory set is evaluated with a preset evaluation index, and the elements in the candidate reversing-into-garage trajectory set are sorted from best to worst according to the evaluation results. The optimal entry trajectory that meets the target conditions is selected from the candidate reversing-into-garage trajectory set as the target path. For example, the optimal solution set can be determined based on the shortest path search comparison, and the elements in the candidate reversing-into-garage trajectory set are sorted from short to long accordingly. The parking spaces and the corresponding reversing-into-garage trajectories can be displayed on the driver's interactive interface in ascending order of distance. The candidate with the highest distance ranking can be selected from the optimal solution set as the target path. In the case where there are other constraints, the candidate with the highest distance ranking that meets other constraints is selected as the target path.

[0032] The main innovation of the present invention is to determine the optimal loading coordinates and path by only using the grid map and the coordinates of the excavator. It ensures that in any complex scene, as long as the complex scene of reversing into the warehouse can be realized manually, the automatic driving mine car can also be solved, avoiding the problem of failure of solving often caused by manual determination of the loading position, multi-dimensional solution, and better robustness of the algorithm. The present invention is based on a method that only relies on the longitude and latitude coordinates of the excavator to calculate the optimal loading position and loading path. The core of this method is different from other methods in that it does not rely on the excavator driver to provide the loading coordinates and vehicle orientation, and can calculate the optimal loading position according to the position of the excavator and the surrounding environment, and ensure that the loading position is optimal. The present invention can calculate the optimal loading position and orientation, and ensure that the reversing into the warehouse path is optimal, which can maximize the efficiency of reversing into the warehouse, and has good adaptability for complex scenes, especially for environments where the excavator cannot determine the loading position and orientation angle by posture.

[0033] The above content only describes the preferred implementation mode of the present invention, and does not limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solution of the present invention by ordinary technicians in this field should fall within the protection scope determined by the claims of the present invention.

Claims

1. A method for planning the shovel position of an autonomous driving mine vehicle, characterized in that: The following steps are involved: S1. Determine the parking area for the mine car: obtain the location of the loading point, determine the theoretical loading area around the loading point, and remove the obstacle area from the theoretical loading area to obtain the loading area; S2, generating a candidate parking space set step: generating a plurality of parking spaces in the loading area according to set rules to form a candidate parking space set, wherein the parking spaces are represented by parking space coordinates and orientation angles; S3, generating a candidate reverse parking trajectory set step: generating a reverse parking trajectory for each parking space in the candidate parking space set to form a candidate reverse parking trajectory set; S4, step of selecting a trajectory that meets the requirements as the target path: evaluating each reversing-into-garaging trajectory in the candidate reversing-into-garaging trajectory set with a preset evaluation index, and selecting the reversing-into-garaging trajectory whose evaluation result meets the preset requirements as the target path.

2. The method for planning the shovel position of an autonomous driving mine vehicle according to claim 1, characterized in that: In the step S1 of determining the area where the mine car can be parked, the theoretical loading area is determined based on the maximum operating radius and the minimum operating radius of the operating vehicle such as the excavator and the size of the mine car compartment.

3. The method for planning the shovel position of an autonomous driving mine vehicle according to claim 1, characterized in that: The maximum operating radius, the minimum operating radius and the size of the mine car compartment are determined based on the obtained loading equipment and the type of mine car.

4. The method for planning the shovel position of an autonomous driving mine vehicle according to claim 1, characterized in that: In the step S1 of determining the mine car parking area, a grid map of the surrounding environment of the loading point is obtained, the grid map and the theoretical loading area are superimposed, and then the obstacle area in the grid map is deleted from the theoretical loading area to determine the loading area.

5. The method for planning the shovel position of an autonomous driving mine vehicle according to claim 1, characterized in that: In the step S2 of generating a candidate parking space set, a number of parking spaces are determined according to a preset rule to obtain a parking space set; and parking spaces whose projection area of ​​the mining car exceeds the loading area are eliminated to obtain a candidate parking space set.

6. The method for planning the shovel position of an autonomous driving mine vehicle according to claim 5, characterized in that: In the step of eliminating parking spaces where the projection area of ​​the mine cart exceeds the loading area, a grid area is generated for each parking space based on the size, center point and orientation angle of the mine cart, and it is determined whether the grid area is within the loading area; Only the parking spaces that are completely within the loading area are retained as the candidate parking space set.

7. The method for planning the shovel position of an autonomous driving mine vehicle according to claim 5, characterized in that: In the step of determining a number of parking spaces according to preset rules, a number of mine car center points are selected in the loading area, and then a number of vehicle orientation angles are generated at each mine car center point. The mine car center points and orientation angles are combined to obtain a number of parking spaces to form a parking space set.

8. The method for planning the shovel position of an autonomous driving mine vehicle according to claim 7, characterized in that: In the step of selecting several center points of the mine cart, an optimal loading line is determined in the loading area, and several coordinate points are selected on the optimal loading line as the center points of the mine cart.

9. The method for planning the shovel position of an autonomous driving mine vehicle according to claim 1, characterized in that: In the step S4 of selecting a trajectory that meets the requirements as the target path, the elements in the candidate reverse parking trajectory set are sorted from best to worst according to the evaluation result, and the optimal parking trajectory that meets the target conditions is selected from the candidate reverse parking trajectory set as the target path.

10. The method for planning the shovel position of an autonomous driving mine vehicle according to claim 9, characterized in that: The step S4 selects a trajectory that meets the requirements as the target path and determines the evaluation result based on the shortest path search comparison.

Images