High-performance parking path planning method for narrow pit-shoveling scene of mining area

By using rasterized environment maps and puffed rectangle coverage technology in narrow spade scenes, reasonable mine car parking paths are enumerated and screened, and the completeness and efficiency of path planning in the existing technology is solved, and the more Efficient planning of autonomous driving minecar paths.

CN118936479APending Publication Date: 2024-11-12TAGE IDRIVER TECHNOLOGY CO LTD
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411284761.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-13
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The existing mine car parking path planning method has problems such as insufficient algorithm completeness, easy to fall into dimensional disasters and high initial solution dependence in narrow shovel scenarios, resulting in difficult path planning and slow calculation speed, which cannot be fully applicable to automatic Parking path planning tasks for driving a mine car.

Method used

By importing a rasterized environment map, identify equally spaced seed points and expanding the puffed rectangle to cover the blank area, enumerate the pointed points to form an alternative path, eliminate the collision path, and finally select the shortest reasonable path as output, avoiding legal path omissions and Dimensional disasters improve calculation speed and algorithm completeness.

Benefits of technology

The mine car parking path planning with more algorithmic completeness and computing efficiency in narrow shovel scenarios is realized, which avoids omissions and solutions of path planning, and expands the application scope of the method.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118936479A_ABST
    Figure CN118936479A_ABST
Patent Text Reader

Abstract

The invention belongs to the field of mine car path planning, and particularly relates to a high-performance parking path planning method for a narrow pit digging scene of a mining area. According to the scheme, firstly, a rasterized environment map and position and posture information of a starting point and an end point are imported, a path from the starting point to the end point is determined from a non-obstacle-occupied area in the map, and seed points at equal intervals on the path and the number M of the seed points are obtained; a preset module A is used for determining a set S composed of M puffing rectangles corresponding to the M seed points according to the position of each seed point and the map, and the total area Sare of all the puffing rectangles in the set S is calculated; according to the scheme, the problems existing in a classic category technical method of an automatic driving mine car parking path planning task are solved, so that the scheme is more suitable for a narrow pit-shoveling parking path planning task in the automatic driving mine car parking path planning task than the classic category technical method.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This solution belongs to the field of mine vehicle path planning, and specifically involves a high-performance parking path planning method for narrow shovel pit scenarios in mining areas. Background Art

[0002] The parking path planning function is a must-have capability for autonomous mine trucks in open-pit mines when driving in narrow shovel pits. Narrow shovel pits refer to the blasting pile at the foot of the open-pit mine or near the excavator, where the mine truck can pass through the excavator's continuous excavation. Due to the insufficient excavation width, such a passable area is narrow, which means that the driving process is difficult. Since the truck needs to reverse to the deep shovel pit in order to carry the gravel or minerals placed by the excavator's electric shovel, the mine truck path planning problem in the narrow shovel pit scenario is a recognized parking path planning task.

[0003] At present, the parking path planning methods suitable for narrow shovel pit scenarios include sampling methods, search methods and numerical optimization methods. The sampling method lacks algorithmic completeness and often cannot find any legal path or the optimal driving path because the sampling resolution is not high enough; the search method can ensure the optimality of the path, but it is easy to fall into the dimensionality disaster, resulting in difficulty in path planning; the numerical optimization method is too dependent on the initial solution, and the numerical optimization is relatively slow, so the scope of application is limited.

[0004] Currently, the classic technical methods applicable to the parking path planning task of autonomous driving mine cars cannot be fully applied to the parking path planning task of narrow shovel pits. Summary of the invention

[0005] The purpose of this solution is to provide a high-performance parking path planning method for narrow shovel pit scenarios in mining areas, in order to solve the problem that the classic category technical methods for autonomous driving mine car parking path planning tasks cannot be fully applied to narrow shovel pit parking path planning tasks.

[0006] In order to achieve the above object, this solution provides a high-performance parking path planning method for narrow shovel pit scenarios in mining areas, which is characterized by comprising the following steps:

[0007] S10: import the rasterized environment map map and the starting point and end point pose information, determine a path from the starting point to the end point in the non-obstacle occupied area in the map, and obtain the equally spaced seed points on the path and the number of seed points M;

[0008] S20: using a preset module A to determine a set S consisting of M puffed rectangles corresponding to the M seed points according to the position of each seed point and the map, and calculating the total area S_are of all puffed rectangles in the set S;

[0009] S30: receiving the user-specified parameter as the total number of cusps N, and then determining the number of cusps corresponding to the m-th puffed rectangle from N according to the area ratio of the puffed rectangle in S_are, and determining a candidate cusp set SS with a uniformly distributed matrix structure according to the length-width ratio of the m-th puffed rectangle, and the number of candidate cusps in the set SS is closest to N m ;

[0010] S40: All candidate sharp points in the set SS are taken as turning points, so as to calculate the reversing path between the starting point and the end point posture passing through the turning point as the alternative path, and all the alternative path sets are recorded as SSS. Then, the alternative paths in the set SSS where the vehicle body footprint collides with the obstacles in the map are removed, and finally the shortest alternative path in the set SSS is selected as the final output result.

[0011] The principle of this scheme is: this scheme divides the path between the starting point and the end point posture at equal intervals to determine the seed point, expands the largest puffed rectangle that does not overlap with obstacles with the seed point as the geometric center, and allows all puffed rectangles to cover all blank areas in the map as much as possible. Then, find points that may become cusps in the blank area, enumerate the positions and orientations of the cusps, and smoothly connect them with the starting point and the end point posture (i.e., position + posture) on both sides, so as to form a series of alternative complete paths (i.e., alternative paths), and then eliminate the paths involving collisions, and further select the most reasonable path from the legal paths according to a certain indicator.

[0012] The principle of this scheme is as follows: (1) When the puffed rectangle covers the blank area, two adjacent puffed rectangles overlap each other under the influence of the map terrain. This allows all puffed rectangles to cover all blank areas in the map as much as possible, so that the collected sharp points can cover the blank areas more comprehensively and accurately, thereby making the enumerated alternative paths more comprehensive, and then eliminating the paths involving collisions in the alternative paths to obtain the final legal paths, avoiding the omission of legal paths, making this scheme more algorithmically complete. (2) The enumeration method used in this scheme is limited by the map and the distance between path segmentation, and will not cause dimensionality disasters, thereby avoiding the difficulty of path planning solution. (3) The enumeration and screening method of this scheme does not rely on the initial solution, and the calculation speed is more controllable, so it has a wider range of applications.

[0013] To sum up: this solution solves the problems existing in the classical category technical methods of the autonomous driving mine car parking path planning task, making this solution more suitable for the narrow shovel pit parking path planning task than the classical category technical methods in the autonomous driving mine car parking path planning task.

[0014] Furthermore, when obtaining the number M of seed points in S10, a preset spacing L is used, the starting point is taken as the first seed point, and the positions of other seed points and the number M of seed points on the path are determined by the length of the path and the spacing L.

[0015] When using the same number of seed points to segment maps with large size differences, the error calculated for larger maps will be greater. In this case, using a fixed spacing L to segment the path (the path between the starting point and the end point) for larger maps can segment the path more accurately and provide an accurate data basis for subsequent enumeration calculations.

[0016] Furthermore, when the number M of seed points is obtained in S10, the spacing between the seed points is determined by the length of the path and the preset number M, and then the position of each seed point is determined by the spacing.

[0017] When the size of the maps to be segmented is small, the segmentation method using a fixed number of seed points can accurately control the number of enumerations, which is beneficial for controlling the calculation speed of the path and saving computing power.

[0018] Furthermore, in step S20, a specified step length Δs is preset in module A. For the mth seed point Q m , with Q m A rectangle with a length and width of 0 is established for the geometric center. It is expanded outward along the four sides of the rectangle in a clockwise or counterclockwise direction by Δs until the expanded part overlaps with the puffing obstacle. Then the expansion attempt in that direction is stopped. After the expansion is terminated in all four directions, Q is finally obtained. m The corresponding puffed rectangle S m .

[0019] This method can quickly and accurately expand the expanded rectangle with the seed point as the geometric center, further improving the calculation speed and saving computing power while achieving accurate coverage of the blank areas in the map.

[0020] Furthermore, the initial value of Δs in module A is 0.1 meters. If the user specifies a step parameter, Δs is changed to the step parameter specified by the user.

[0021] Furthermore, in the step S30, it is recommended to the user to select an integer between 500 and 1200 as the value of the total number of cusps N.

[0022] By using the method of user-specified step size and total number of cusps, the scheme can be applied to a wider range through user judgment.

[0023] Furthermore, in step S30, the number of candidate cusps in set SS is closest to N m and not less than N m .

[0024] This method can control the number of cusps to a minimum, thereby further saving computing power.

[0025] Further, in the step S40, determining the set SSS includes the following steps:

[0026] S41: taking the straight line between the starting point and the end point as the start-end line;

[0027] S42: Take a candidate cusp point from the set SS as point P, and determine whether the position of point P is on the start-end line. If:

[0028] S42a: The position of point P is on the start-end line, and the process returns to S42;

[0029] S42b: The position of point P is not on the line connecting the start and the end. Draw a perpendicular line from point P to the line connecting the start and the end. The foot of the perpendicular is recorded as Q. The attitude direction of point P is required to be set as a vector direction;

[0030] S43: Connect the shortest path curve from the starting point to point P as the first curve, and the posture direction of the first curve when it connects to point P is along direction; in a similar way, the shortest path curve from point P to the end point is connected as the second curve, and the attitude direction of the second curve at point P is along direction, and combine the first curve and the second curve to obtain an alternative path, and put the alternative path into a set SSS;

[0031] S44: Repeat steps S42 to S43 until all candidate cusps in SS are exhausted, and output set SSS.

[0032] Furthermore, in the step S40, when selecting the shortest candidate path in the set SSS, the set SSS is first traversed, and then the length of each candidate path is calculated, and finally the shortest candidate path is compared and obtained as the final output result.

[0033] By enumerating the cusps and calculating the alternative paths corresponding to the cusps, and then screening the alternative paths one by one, the legal paths can be accurately screened out, and the final output results can be accurately screened out without missing any paths.

[0034] Furthermore, in the step S43, when calculating the shortest path curve, R X is the radius, the R X =R min γ, where R min is the minimum turning radius of the mine cart, γ is a parameter greater than 1, and γ is 1.0, 1.5 and 2.0 respectively, to obtain the first curve set D1 and the second curve set D2. The alternative path is a combination of any first curve in D1 and any second curve in D2.

[0035] The minecart can pass through the first curve, but when the second curve collides with an obstacle, the path curve combined with the first curve and the second curve will be excluded. By controlling the radius of the path curve, the first curve and the second curve are calculated separately, and then different first curves and second curves are combined. The path curves that the minecart can pass through corresponding to the sharp point can be retained as much as possible, thereby retaining the alternative paths passing through the sharp point as much as possible, further avoiding path omissions. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is a schematic diagram related to path in the embodiment of this scheme.

[0037] Figure 2 This is a schematic diagram of the puffed rectangle in the embodiment of this scheme.

[0038] Figure 3 This is a schematic diagram of the alternative sharp points in the embodiment of this scheme.

[0039] Figure 4 This is a schematic diagram of the connection between the beginning and the end in this embodiment.

[0040] Figure 5 Schematic diagram of the Dubins curve under different γ parameters in the embodiment of this scheme.

[0041] Figure 6 Schematic diagram of the method for generating a puffed rectangle in module A in the embodiment of this scheme.

[0042] Figure 7 It is the rasterized environment map map and the starting point pose and end point pose information in the embodiment of this solution. DETAILED DESCRIPTION

[0043] The following will clearly and completely describe the concept and technical effects of the present invention in combination with the embodiments, so as to fully understand the purpose, features and effects of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, other embodiments obtained by those skilled in the art without creative work are all within the scope of protection of the present invention:

[0044] The embodiment is basically as shown in the attached Figure 1-7 As shown:

[0045] The high-performance parking path planning method for the narrow shovel pit scenario in the mining area includes the following steps:

[0046] S11: Import the rasterized environment map and the starting and ending pose information (see attached Figure 7 shown);

[0047] S12: Use the A-star search algorithm to search for a path connecting the starting point and the end point in the non-obstacle occupied area in the map, which is called path (see the attached Figure 1 shown);

[0048] S13: Select M equally spaced sampling points on the path by presetting the spacing L or the number M, which are called seed points, and determine the position of each seed point.

[0049] S21: For each seed point, generate a puffed rectangle containing the seed point according to the method given in module A, and determine M puffed rectangles accordingly. Record these M puffed rectangles in a set S (see attached Figure 2 shown).

[0050] Among them, the input list of module A includes a seed point (called seed point V) and map, and outputs a puffed rectangle.

[0051] The seed point V can be regarded as a rectangle with a length of 0 and a width of 0. It is expanded outward along the four sides of the rectangle in a clockwise or counterclockwise direction with a unit step length Δs (if the user specifies the step length parameter, Δs is changed to the step length parameter specified by the user) until the expanded part overlaps with the puffed obstacle, then the expansion attempt in this direction is stopped.

[0052] The algorithm diagram of module A to generate the puffed rectangular area is shown below:

[0053]

[0054]

[0055] The parameter appearing in line 9 of the algorithm is used to limit the maximum allowable expansion size of the puffed rectangle, so as to avoid wasting calculation time due to generating a puffed rectangle that is too large.

[0056] As attached Figure 6 As shown in , each local incremental rectangular area will be explored from small to large to see if it collides with the puffed obstacle: if no collision occurs, the expanded part will be merged into the rectangle, otherwise the expanded part will be abandoned and no further attempts will be made to expand in that direction. After the expansion is terminated in all four directions, the final puffed rectangle is obtained. Figure 6 It can be seen that the seed point V is not necessarily the geometric center of the finally determined puffed rectangle.

[0057] S22: Extract M expanded rectangles from the set S. Assume that the areas of each rectangle are area1, area2, …, area M , now use the following formula to solve the sum of the areas of all rectangles area_sum;

[0058]

[0059] Based on this, we can calculate the proportion of the total area occupied by each puffed rectangle.

[0060] S31: Now set a total of N candidate cusp quotas (N is a parameter specified by the user, generally an integer between 500-1200), then the number of candidate cusps allocated to the i-th puffed rectangle is num_cusps i It can be estimated as follows:

[0061]

[0062] The round() function is used to round off. By traversing i=1,2,…,N, we can get the number of cusps assigned to each puffed rectangle: num_cusps i .

[0063] S32: For the i-th puffed rectangle, sample a rows and b columns of candidate cusps uniformly inside it according to its length-to-width ratio, so that the value of a multiplied by b is as close as possible to num_cusps i Specifically, let's assume that the length of the i-th puffed rectangle is L and the width is W, then the values ​​of a and b are determined by the following formula:

[0064]

[0065] After determining the parameters a and b, arrange a×b candidate cusps along the x-axis and y-axis in an arithmetic progression on the expanded rectangle, as shown in the following figure. Figure 3 As shown, the puffed rectangle Sm is uniformly sampled into 5 rows and 4 columns of candidate cusps (ie, a=5 and b=4) according to the ratio of length to width.

[0066] By traversing i=1, 2, ..., N, we can obtain the candidate cusps sampled inside each puffed rectangle. Now all the candidate cusps are recorded in the set SS.

[0067] S41: taking the straight line between the starting point and the end point as the start-end line;

[0068] S42: Take a candidate cusp point from the set SS as point P, and then determine on which side of the line connecting the start and end positions point P falls, if:

[0069] S42a: If point P falls on the start-end line, directly eliminate the point P;

[0070] S42b: If point P falls on one side of the line connecting the start and the end, then draw a perpendicular line from point P to the line connecting the start and the end, and record the foot of the perpendicular as Q. The attitude direction of point P is required to be set as the vector Direction (if attached) Figure 4 shown).

[0071] S43: The radius from the starting point to point P is R x Dubins curve, and the attitude direction of the curve when it connects to point P is along Direction. Here R x The value is

[0072] R x =R min γ (4)

[0073] Where R min It refers to the minimum turning radius of the vehicle. γ is a parameter greater than 1. It is required that γ takes three values ​​of 1.0, 1.5 and 2.0, corresponding to the three connecting starting position and Dubins curve of direction (see attached Figure 5 shown).

[0074] In a similar way, the Dubins curves under three parameters are connected from point P to the end point, and the attitude direction of the curve at point P is ensured to be along Direction (if attached) Figure 5 shown).

[0075] Due to the combination of parameters, there are 3 possibilities for the first half of the curve, and 3 independent possibilities for the second half of the curve. This means that if we try to connect the start and end positions with P as the cusp, we will combine 9 complete parking paths that are not exactly the same. We then store these 9 complete paths.

[0076] S44: Repeat the above steps (ie, steps S42 to S43), perform the above calculation and save for each candidate cusp P in the SS set, and then obtain a complete path set of candidate paths. It is advisable to set the set storing all candidate paths as SSS.

[0077] S45: Extract each element in the SSS set (i.e., a specific candidate path) respectively. If the vehicle footprint corresponding to the path collides with the map, then directly remove the path, otherwise calculate the length of the path. By traversing the lengths of all paths in the SSS, we select the path with the shortest length and output it as the final output result of the entire parking path planning system.

[0078] The above is only an embodiment of the present invention, and the common knowledge such as the known specific structure and characteristics in the scheme is not described in detail here. It should be pointed out that for those skilled in the art, several deformations and improvements can be made without departing from the structure of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A high-performance parking path planning method for narrow shovel pit scenarios in mining areas. It is characterized in that The following steps are involved: S10: import the rasterized environment map map and the starting point and end point pose information, determine a path from the starting point to the end point in the non-obstacle occupied area in the map, and obtain the equally spaced seed points on the path and the number of seed points M; S20: using a preset module A to determine a set S consisting of M puffed rectangles corresponding to the M seed points according to the position of each seed point and the map, and calculating the total area S_are of all puffed rectangles in the set S; S30: receiving a user-specified parameter as the total number of cusps N, and then determining the number of cusps N corresponding to the mth puffed rectangle from N according to the area ratio of the puffed rectangle in S_are m According to the length-width ratio of the mth puffed rectangle, a candidate point set SS with a uniform distribution matrix structure is determined, and the number of candidate points in the set SS is closest to N. m ; S40: All candidate sharp points in the set SS are taken as turning points, so as to calculate the reversing path between the starting point and the end point posture passing through the turning point as the alternative path, and all the alternative path sets are recorded as SSS. Then, the alternative paths in the set SSS where the vehicle body footprint collides with the obstacles in the map are removed, and finally the shortest alternative path in the set SSS is selected as the final output result.

2. According to claim 1, the high-performance parking path planning method for narrow shovel pit scenarios in mining areas, Features: When obtaining the number M of seed points in S10, a preset spacing L is used, the starting point is taken as the first seed point, and the positions of other seed points and the number M of seed points on the path are determined by the length of the path and the spacing L.

3. According to claim 1, the high-performance parking path planning method for narrow shovel pit scenarios in mining areas, Features: When the number M of seed points is obtained in S10, the distance between the seed points is determined by the length of the path and the preset number M, and then the position of each seed point is determined by the distance.

4. According to any one of claims 2 or 3, the high-performance parking path planning method for narrow shovel pit scenarios in mining areas, Features: In step S20, a specified step length Δs is preset in module A. For the mth seed point Q m , with Q m A rectangle with a length and width of 0 is established for the geometric center. It is expanded outward along the four sides of the rectangle in a clockwise or counterclockwise direction by Δs until the expanded part overlaps with the puffing obstacle. Then the expansion attempt in that direction is stopped. After the expansion is terminated in all four directions, Q is finally obtained. m The corresponding puffed rectangle S m .

5. According to claim 4, the high-performance parking path planning method for narrow shovel pit scenarios in mining areas, Features: The initial value of Δs in module A is 0.1 meters. If the user specifies a step parameter, Δs is changed to the step parameter specified by the user.

6. According to claim 5, the high-performance parking path planning method for narrow shovel pit scenarios in mining areas, Features: In the step S30, it is recommended to the user to select an integer between 500 and 1200 as the value of the total number of cusps N.

7. According to claim 6, the high-performance parking path planning method for narrow shovel pit scenarios in mining areas, Features: In the step S30, the number of candidate cusps of the set SS is closest to N m and not less than N m .

8. According to claim 7, the high-performance parking path planning method for narrow shovel pit scenarios in mining areas, Features: In the step S40, determining the set SSS includes the following steps: S41: taking the straight line between the starting point and the end point as the start-end line; S42: Take a candidate cusp point from the set SS as point P, and determine whether the position of point P is on the start-end line. If: S42a: The position of point P is on the start-end line, and the process returns to S42; S42b: The position of point P is not on the line connecting the start and the end. Draw a perpendicular line from point P to the line connecting the start and the end. The foot of the perpendicular is recorded as Q. The attitude direction of point P is required to be set as a vector direction; S43: Connect the shortest path curve from the starting point to point P as the first curve, and the posture direction of the first curve when it connects to point P is along direction; in a similar way, the shortest path curve from point P to the end point is connected as the second curve, and the attitude direction of the second curve at point P is along direction, and combine the first curve and the second curve to obtain an alternative path, and put the alternative path into a set SSS; S44: Repeat steps S42 to S43 until all candidate cusps in SS are exhausted, and output set SSS.

9. The high-performance parking path planning method for narrow shovel pit scenarios in mining areas according to claim 8, Features: In the step S40, when selecting the shortest candidate path in the set SSS, the set SSS is first traversed, and then the length of each candidate path is calculated, and finally the shortest candidate path is compared and obtained as the final output result.

10. The high-performance parking path planning method for narrow shovel pit scenarios in mining areas according to claim 9, Features: In the step S43, when calculating the shortest path curve, R X is the radius, the R X =R min γ, where R min is the minimum turning radius of the mine cart, γ is a parameter greater than 1, and γ is 1.0, 1.5 and 2.0 respectively, to obtain the first curve set D1 and the second curve set D2. The alternative path is a combination of any first curve in D1 and any second curve in D2.

Citation Information

Cited By

  • Parking path generation method and device

    CN120096551A