Multi-target path planning method based on two-dimensional grid map

Abstract

The present application relates to the technical field of logistics robot path planning, and particularly relates to a multi-target path planning method based on a two-dimensional grid map. The method comprises the following steps: obtaining a two-dimensional grid map of a known environment, and converting the grid map into a cost map; given a starting point and a plurality of target points of a mobile robot, the starting point and the target points are combined in order to obtain a plurality of local areas; an MTJPS algorithm is used to find a path from the starting point to the target point in each local area, and the paths of the local areas are combined into a complete path; a turning point in the path is selected to obtain adjacent control points, and the turning point in the path is optimized by using a B-spline curve to obtain an optimized path; and the optimized path is combined with the original path to obtain a final path. The method can solve the problem of multi-target path planning, and obtain a more optimal global path, which is more suitable for path planning of a logistics robot.

Description

Technical Field

[0001] This invention relates to the field of logistics robot path planning technology, specifically to a multi-objective path planning method based on a two-dimensional grid map. Background Technology

[0002] Path planning, which involves planning a traversable path from a starting point to a target point, is a crucial component of logistics robot navigation and shows great promise in scenarios such as material warehouses, production workshops, and hospital sterilization supply centers. For example, the multimodal, multi-objective path planning algorithm disclosed in Chinese Patent 202210444997.4 stores target value information from a map into an adjacency matrix; then it defines and initializes variables for the multimodal, multi-objective algorithm; finally, it uses the multimodal, multi-objective algorithm, the adjacency matrix, and the initialized variables for path planning. However, in practical applications, scenarios are often complex, involving obstacles, pedestrians, and numerous unknown risks, with many factors being uncontrollable. Single-objective path planning cannot meet the needs of real-world applications. Therefore, a multi-objective path planning algorithm more suitable for practical applications is needed to avoid planning paths that traverse obstacles or work areas. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a multi-objective path planning method based on two-dimensional grid maps, which can plan a travel path that is more in line with the actual application scenario.

[0004] The technical solution of this invention is as follows:

[0005] A multi-objective path planning method based on a two-dimensional grid map includes the following steps:

[0006] S1. Obtain a known environmental cost map: This includes the following steps:

[0007] S11. Obtain a two-dimensional raster map of the known environment;

[0008] S12. Convert the two-dimensional raster map into a high-resolution cost map with the same resolution as the two-dimensional raster map.

[0009] S13. Convert the high-resolution cost map to a low-resolution cost map, with the same map size but different resolution;

[0010] S2, MTJPS planning path: includes the following steps:

[0011] S21: Given the starting point of the mobile robot and multiple target points;

[0012] S22: Combine the starting point and multiple target points in pairs sequentially to divide them into multiple local regions;

[0013] S23: Use the MTJPS algorithm to find a path from the starting point to the target point in each local region;

[0014] S24: Combine the paths of the various local regions into a complete path;

[0015] S3. Inflection Point Path Optimization: This includes the following steps:

[0016] S31: Select the inflection points in the complete path;

[0017] S32: Obtain adjacent control points based on inflection points.

[0018] S33: Optimize inflection points in the path using B-spline curves;

[0019] S34: Obtain the optimized path, and combine it with the original path to obtain the final path.

[0020] Preferably, in step S11, the two-dimensional raster map M, the high-resolution cost map CM, and the low-resolution cost map CM′ are respectively represented as:

[0021] M = {m n =(x n y n )|m n ∈[0,1]}

[0022] CM = {cm n =(x n y n )|cm n ∈[0...255]}

[0023] CM′={cm n′ =(x n′ y n′ )|cm n′ ∈[0...255]}

[0024] In the formula: n=[0...(H×W)], n′=[0...(H′×W′)], representing the number of map pixels;

[0025] H, W, H′, and W′ represent the length and width of the map, respectively. The resolution of the 2D raster map and the high-resolution cost map can be modified and is set to 0.02m; the resolution of the low-resolution cost map is... Where N is a variable parameter.

[0026] Preferably, in step S13, the conversion from a high-resolution cost map to a low-resolution cost map includes the following specific steps:

[0027] S131. Calculate the length H′ and width W′ of CM′, and obtain:

[0028] H′=H÷N,W′=W÷N

[0029] S132. Calculate the cost of CM′, and obtain:

[0030] Cost = f sum (CM), CM′=f avg (Cost)

[0031] In the formula: f sum This represents the accumulation function, where Cost represents the accumulated cost, f avg This represents the average function, yielding a new cost value.

[0032] Preferably, in step S21, the starting point is the pose of the mobile robot in the grid map, including position coordinates and attitude angles; the multiple target points consist of several sub-target points and a target endpoint, including the position coordinates of each target point; the starting point and the multiple target points are represented as follows:

[0033] P = {p t =(x t y t |t=[0...T]}

[0034] In the formula: T represents the starting point and the number of multiple target points.

[0035] Preferably, in step S22, the local region is the area between only two target points, which is a subset of the cost map. The cost map includes a high-resolution cost map and a low-resolution cost map; then the local region M′ is represented as:

[0036] M′={m′ ij =(x ij y ij )|m′ min <m′ ij <m′ max}

[0037] In the formula: i = [0...T-1], representing the number of local regions;

[0038] j = [0...(x t -x t+1 )×(y t -y t+1 ] represents the number of pixels in a local region;

[0039] Each local region is obtained from two adjacent target points, including the following specific steps:

[0040] S221. Calculate the coordinates m′ of the minimum point in the local region M′. min (xmin y min ):

[0041]

[0042] S222. Calculate the maximum point coordinates m′ in the local region M′. max (x max y max ):

[0043]

[0044] In the formula: These represent functions that take the minimum values ​​along the X and Y axes, respectively.

[0045] These represent functions that take the maximum value along the X-axis and Y-axis, respectively.

[0046] Preferably, in step S23, the MTJPS algorithm is a modification of the JPS algorithm, including the search direction, search rules, and heuristic function, wherein:

[0047] Search direction: the first quadrant, second quadrant, third quadrant, and fourth quadrant in the Cartesian coordinate system. The actual search direction is determined by the relative position of the starting point and multiple target points.

[0048] Search rules: There are two types of rules:

[0049] Rule 1: Search for jump points along the X-axis. After encountering obstacles or local area boundaries, increment along the Y-axis. The search process terminates under the following conditions:

[0050] a. An obstacle or local boundary is encountered in the Y-axis direction; b. The jump count is zero; c. The target point is found;

[0051] Rule 2: The search for jump points is along the Y-axis. After encountering an obstacle or local boundary, the search increments along the X-axis. The termination conditions for the search process include:

[0052] a. An obstacle or local boundary is encountered in the X-axis direction; b. The jump count is zero; c. The target point is found;

[0053] Heuristic function: The heuristic function f(a) represents the total cost from the starting point to the target point, including the following specific steps:

[0054] S231. Calculate the length g(a) of the search path from the starting point to the current point a:

[0055] g(a)=f sum (p0, a)

[0056] S232. Calculate the cost h(a) from the current point to the target point:

[0057] h(a)=f sqrt ((a x -e x ) 2 -(a y -e y ) 2 )

[0058] S233. Calculate the obstacle cost o(a) from the current point to the target point:

[0059] o(a)=f ox (a x e x )+f 0y (a y e y )

[0060]

[0061] S234. Calculate the total cost f(a) from the starting point to the target point:

[0062] f(a) = g(a) + h(a) + o(a)

[0063] In the formula: a∈M′, e∈P, a x a y e x and e y Let a and e represent the coordinates of points a and e, respectively.

[0064] p0 represents the starting point, f sum f represents the cumulative search path length from the starting point to the current point. sqrt f represents the square root function. ox f 0y These represent the obstacle rate functions calculated from the current point to the target point along the X and Y axes, respectively. o f all These represent the number of obstacle grids and the total number of grids, respectively.

[0065] Preferably, in step S24, the complete path includes sub-paths of each local region, which are recombined based on the path point positions; the complete path PATH is represented as:

[0066] PATH = {path i′j′ =(x j ′,y j ′)|i′=[0...T-1]}

[0067] In the formula: i′ represents the number of sub-paths, and j′ represents the number of path points in the i′-th sub-path;

[0068] The calculation of the complete path PATH involves the following specific steps:

[0069] S241. Calculate the path in the local region. j′ :

[0070] path j′ =f mtjps (p t p t+1 )

[0071] S242, Calculate the total path. i′j′ :

[0072] path i′j′ =f CM (f append (path j′ ))

[0073] In the formula: f append Indicates path concatenation, f mtjps This refers to the MTJPS function, used to generate p t and p t+1 Subpaths between, f CM Used to restore waypoints to a high-resolution cost map (CM).

[0074] Preferably, in step S31, the inflection point is a point that forms a right angle with an adjacent path point in the path, and the inflection point InfP is represented as:

[0075] InfP = {infp} k =(x k y k )|0≤k<K]}

[0076] In the formula: Indicates the number of path points;

[0077] The determination of the inflection point InfP includes the following specific steps:

[0078] S311, Calculate path j′ (p t-1 p t ) angle θ i :

[0079] θ i =arctan2(p t-1 p t )

[0080] S312, Calculate path i′+1 (p t p t+1 ) angle θ i+1 :

[0081] θ i+1 =arctan2(p t p t+1 )

[0082] S313. Determine if the current point is an inflection point:

[0083] θ=f abs (θ i -θ i+1 )

[0084]

[0085] In the formula: arctan2 is the inverse trigonometric function, f abs f is an absolute value function inflection This is the inflection point determination function.

[0086] Preferably, in step S32, the control point is generated from the inflection point, and the control point CP is represented as:

[0087] CP = {cp k′ =(x k′ , yk′ )|0≤k′ <K′]}

[0088] In the formula: K′ represents the number of control points, which is a variable parameter;

[0089] The calculation of control points (CP) includes the following specific steps:

[0090] S321, Calculate path i′ (p t-1 p t Control point coordinates (x) in the direction k′ y k′ ):

[0091]

[0092] S322, Calculate path i′+1 (p t p t+1 Control point coordinates (x) in the direction k′+1 y k′+1 ):

[0093]

[0094] S323. Calculate the total control points:

[0095] CP = f append (cp k′ )

[0096] In the formula: f represents the function that calculates the X-axis and Y-axis coordinates of the control point, respectively. append This indicates the splicing of control points, where d represents the unit distance from the control point to the inflection point, and is a variable parameter.

[0097] Preferably, in step S24, the final path includes the inflection point optimized path and the original path, and the final path PATH final Represented as:

[0098] PATH final ={path i″ =(x i″ y i″ )|i″>0}

[0099] In the formula: i″ represents the number of path points;

[0100] Final path PATH final The calculation includes the following specific steps:

[0101] S241. Calculate the inflection point optimization path:

[0102] path optimize =f bspine (cp k′ )

[0103] S242. Calculate the total path:

[0104] path i″ =f combin (pat i′j′ path optimize )

[0105] In the formula: f bspine Let f represent the B-spline curve optimization function. combin This function represents the recombination of the original path and the optimized path at the inflection point.

[0106] Compared with the prior art, the present invention has the following advantages:

[0107] (1) The present invention can plan a multi-objective path planning algorithm that is more suitable for actual application scenarios, generate a passable path that passes through multiple target points, and avoid generating a path that crosses obstacles or work areas.

[0108] (2) The path planned by this invention is more suitable for the actual application of logistics robots compared with the paths planned by other algorithms. Attached Figure Description

[0109] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0110] Figure 1 This is a flowchart illustrating the overall process described in an embodiment of the present invention.

[0111] Figure 2(a) is a two-dimensional grid map according to an embodiment of the present invention.

[0112] Figure 2(b) is a high-resolution cost map according to an embodiment of the present invention.

[0113] Figure 2(c) is a low-resolution cost map according to an embodiment of the present invention.

[0114] Figure 2(d) is a location diagram of the target point described in the embodiment of the present invention.

[0115] Figure 3 This is a flowchart illustrating the MTJPS algorithm planning process as described in an embodiment of the present invention.

[0116] Figure 4 This is a schematic diagram illustrating the selection of control points according to an embodiment of the present invention.

[0117] Figure 5 This is a diagram showing the original path planning result as described in an embodiment of the present invention.

[0118] Figures 6(a)-6(b) These are comparison diagrams of the inflection point path optimization described in the embodiments of the present invention.

[0119] Figure 7 This is a diagram showing the final path planning result according to an embodiment of the present invention. Detailed Implementation

[0120] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0121] Example 1

[0122] like Figure 1As shown, the purpose of this invention is to provide a multi-objective path planning method based on a two-dimensional grid map. The method includes the following steps: S1, obtaining a known environmental cost map; S2, planning the path using MTJPS; S3, optimizing the path at inflection points.

[0123] Step S1: Obtain a two-dimensional raster map of the known environment, as shown in Figure 2(a), and convert the two-dimensional raster map into a cost map.

[0124] The cost map described in this embodiment includes a high-resolution cost map and a low-resolution cost map, as shown in Figures 2(b) and 2(c).

[0125] The high-resolution cost map described in this embodiment is obtained from a two-dimensional raster map, and its resolution is consistent with that of the two-dimensional raster map.

[0126] The low-resolution cost map described in this embodiment is obtained by converting the high-resolution cost map; the map sizes are the same, but the resolutions are different.

[0127] In this embodiment, the two-dimensional raster map, high-resolution cost map, and low-resolution cost map are respectively represented as follows:

[0128] M = {m n =(x n , yn )|m n ∈[0,1]}

[0129] CM = {cm n =(x n y n )|cm n ∈[0...255]}

[0130] CM′={cm n′ =(x n′ y n′ )|cm n′ ∈[0...255]}

[0131] In the formula: n=[0...(H×W)], n′=[0...(H′×W′)], representing the number of map pixels;

[0132] H, W, H′, and W′ represent the length and width of the map, respectively. The resolution of the 2D raster map and the high-resolution cost map can be modified and is set to 0.02m; the resolution of the low-resolution cost map is... Where N is a variable parameter.

[0133] The conversion of a high-resolution cost map to a low-resolution cost map described in this embodiment includes the following specific steps:

[0134] Calculate the length H′ and width W′ of CM′ to obtain:

[0135] H′=H÷N,W′=W÷N

[0136] Calculate the cost of CM′ and obtain:

[0137] Cost = f sum (CM), CM′=f avg (Cost)

[0138] In the formula: f sum This represents the accumulation function, where Cost represents the accumulated cost, f avg This represents the average function, yielding a new cost value.

[0139] S2. MTJPS Path Planning: Given a mobile robot's starting point and multiple target points, as shown in Figure 2(d), these points are sequentially paired to divide the robot into multiple local regions. The MTJPS algorithm is used to find a path from the starting point to the target point in each local region. The planning process is detailed below. Figure 3 Then, the paths of each local area are combined into a complete path.

[0140] The starting point described in this embodiment is the pose of the mobile robot in the grid map, including position coordinates and attitude angles.

[0141] The multiple target points described in this embodiment consist of several sub-target points and a target endpoint, and include the position coordinates of each target point.

[0142] The local region described in this embodiment only includes the area between two target points and is a subset of the cost map.

[0143] The MTJPS algorithm described in this embodiment is a modification of the JPS algorithm, including the search direction, search rules, and heuristic functions.

[0144] The complete path described in this embodiment includes sub-paths for each local region, which are recombined based on the location of the path points.

[0145] In this embodiment, the starting point and multiple target points are represented as P = {p t =(x t y t )|t=[0...T]}, where T represents the starting point and the number of multiple target points.

[0146] In this embodiment, the local region M′={m′ ij =(x ij y ij )|m′ min <m′ ij <m′ max}, where i = [0...T-1], represents the number of local regions, j = [0...(x t -x t+1 )×(y t -y t+1 ] represents the number of pixels in a local region. Each local region is obtained by taking two adjacent target points, and includes the following specific steps:

[0147] Calculate the coordinates m′ of the minimum point in the local region M′. min (x min y min ):

[0148]

[0149] Calculate the coordinates m′ of the maximum point in the local region M′. max (x max y max ):

[0150]

[0151] In the formula: These represent functions that take the minimum values ​​along the X and Y axes, respectively.

[0152] These represent functions that take the maximum value along the X-axis and Y-axis, respectively.

[0153] The MTJPS algorithm described in this embodiment has four search directions: the first quadrant, the second quadrant, the third quadrant, and the fourth quadrant in the Cartesian coordinate system. The actual search direction is determined by the relative position of the starting point and multiple target points.

[0154] The MTJPS algorithm described in this embodiment has two search rules. One is to search for jump points along the X-axis, and after encountering obstacles or local region boundaries, increment along the Y-axis. The termination conditions of the search process include: a) encountering obstacles or local region boundaries in the Y-axis direction; b) the number of jump points is zero; c) the target point is found. The other is to search for jump points along the Y-axis, and after encountering obstacles or local region boundaries, increment along the X-axis. The termination conditions of the search process include: a) encountering obstacles or local region boundaries in the X-axis direction; b) the number of jump points is zero; c) the target point is found.

[0155] In this embodiment, the heuristic function f(a) represents the cost from the starting point to the target point, where g(a) represents the search path length from the starting point to the current point a, h(a) represents the cost from the current point to the target point, and o(a) represents the obstacle cost from the current point to the target point. The calculation process of f(a) is as follows:

[0156] Calculate the path length g(a):

[0157] g(a)=f sum (p0, a)

[0158] Calculate the cost h(a) from the current point to the target point:

[0159] h(a)=f sqrt ((a x -e x ) 2 -(a y -e y ) 2 )

[0160] Calculate the obstacle cost o(a) from the current point to the target point:

[0161] o(a)=f ox (a x e x )+f 0y (a y e y ),

[0162]

[0163] Calculate the total cost f(a):

[0164] f(a) = g(a) + h(a) + o(a)

[0165] In the formula: a∈M′, e∈P, a x a y e x and e y Let p0 represent the coordinates of points a and e respectively, and f represent the starting point. sum f represents the cumulative search path length from the starting point to the current point. sqrt f represents the square root function. ox f 0y These represent the obstacle rate functions calculated from the current point to the target point along the X and Y axes, respectively. o f all These represent the number of obstacle grids and the total number of grids, respectively.

[0166] In this embodiment, the complete path PATH = {path i′j′ =(x j′ y j′ Let |i′=[0...T-1]}, where i′ represents the number of sub-paths and j′ represents the number of path points in the i′-th sub-path. The calculation process is as follows:

[0167] Calculate the local region path j′ :

[0168] path j′= f mtjps (p t ,p t+1 )

[0169] Calculate the total path path i′j′ :

[0170] path i′j′ = f CM (f append (path j′ ))

[0171] In the formula: f append represents path splicing, and f mtjps represents the MTJPS function, which is used to generate the sub-path between p t and p t+1 . f CM is used to restore the path points to the high-resolution cost map CM. For the complete path, see Figure 5 .

[0172] S3. Inflection point path optimization: Select the inflection points in the path to obtain adjacent control points, and use the B-spline curve to optimize the inflection points in the path to obtain the optimized path. After combining with the original path, the final path is obtained.

[0173] In this embodiment, the inflection point is the point where the angle between adjacent path points in the path is a right angle.

[0174] In this embodiment, the control point CP = {cp k′ = (x k′ , y k′ )|0 ≤ k′ < K′}, which is generated by the inflection points, where K′ represents the number of control points and is a variable parameter. As Figure 4 shown, there are four cases for the selection of control points, taking two control points on each side. The calculation process is as follows:

[0175] Calculate the coordinates (x i′ (p t-1 , p t ) of the control point in the direction of path k′ , y k′ ):

[0176]

[0177] Calculate the coordinates (x i′+1 (p t , p t+1 ) of the control point in the direction of path k′+1 , y k′+1 ):

[0178]

[0179] Calculate the total control points:

[0180] CP = f append (cp k′ )

[0181] In the formula: f represents the function that calculates the X-axis and Y-axis coordinates of the control point, respectively. append This indicates the splicing of control points, where d represents the unit distance from the control point to the inflection point, and is a variable parameter.

[0182] The final path PATH described in this embodiment final ={path i″ =(x i″ y i″ Let )|i″>0}, where i″ represents the number of path points, including the original path and the optimized path at inflection points. The calculation process is as follows:

[0183] Calculate the inflection point optimization path:

[0184] path optimize =f bspine (cp k′ )

[0185] Calculate the total path:

[0186] path i″ =f combin (path i′j′ path optimize )

[0187] In the formula: f bspine Let f represent the B-spline curve optimization function. combin This represents the function that recombines the original path and the optimized path at the inflection point. A comparison graph of the optimized path at a single inflection point and the original path is shown below. Figures 6(a)-6(b) As shown.

[0188] The above describes an implementation of the multi-objective path planning method based on a two-dimensional grid map according to the present invention. The present invention obtains a cost map from a two-dimensional grid map and uses the MTJPS multi-objective planning algorithm to plan a passable path passing through multiple objective points. Compared with other path planning algorithms, it is more suitable for handling multi-objective path planning problems, and the generated path is more consistent with practical applications. For the results of the multi-objective path planning, see [link to relevant documentation]. Figure 7 .

[0189] Although the present invention has been described in detail with reference to the accompanying drawings and preferred embodiments, the invention is not limited thereto. Various equivalent modifications or substitutions can be made to the embodiments of the invention by those skilled in the art without departing from the spirit and essence of the invention, and such modifications or substitutions should all be within the scope of the invention. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the invention should also be covered within the protection scope of the invention. Therefore, the protection scope of the invention should be determined by the scope of the claims.

Claims

1. A multi-objective path planning method based on a two-dimensional grid map, characterized in that, Includes the following steps: S1. Obtain a known environmental cost map: This includes the following steps: S11. Obtain a two-dimensional raster map of the known environment; two-dimensional raster map High-resolution cost map and low-resolution cost maps They are represented as follows: In the formula: , , representing the number of pixels in the map; , , , These represent the length and width of the map, respectively. The resolution of the 2D raster map and the high-resolution cost map can be modified and is set to 0.02m; the resolution of the low-resolution cost map is... ,in S12 is a variable parameter; the two-dimensional raster map is converted into a high-resolution cost map with the same resolution as the two-dimensional raster map. S13. Converting a high-resolution cost map to a low-resolution cost map, with the same map size but different resolution; the conversion includes the following specific steps: S131, Calculation length and width ,get: S132, Calculation The value of the cost is obtained as follows: In the formula: This represents an accumulation function. This represents the accumulated cost. Representing the average function, we obtain a new cost value; S2, MTJPS planning path: includes the following steps: S21: Given a mobile robot's starting point and multiple target points; the starting point is the mobile robot's pose in the grid map, including its position coordinates and attitude angles; the multiple target points consist of several sub-target points and a target endpoint, including the position coordinates of each target point; the starting point and multiple target points are represented as follows: In the formula: Indicates the number of starting and ending points; S22: The starting point and target points are sequentially paired to divide the area into multiple local regions. Each local region is a subset of the cost map, which includes both high-resolution and low-resolution cost maps. The local regions are then... Represented as: In the formula: , indicates the number of local regions; , representing the number of pixels in a local region; Each local region is obtained from two adjacent target points, including the following specific steps: S221, Calculate the local region Minimum point coordinates : ， S222, Calculate the local region Maximum point coordinates : ， In the formula: , They represent taking shaft and The minimum value function of the axis; , They represent taking shaft and The maximum value function of the axis; S23: Use the MTJPS algorithm to find a path from the starting point to the target point in each local region; the MTJPS algorithm is a modification of the JPS algorithm, including the search direction, search rules, and heuristic functions, where: Search direction: the first quadrant, second quadrant, third quadrant, and fourth quadrant in the Cartesian coordinate system. The actual search direction is determined by the relative position of the starting point and the target point. Search rules: There are two types of rules: Rule 1: Along The axis search jumps to a point, and after encountering an obstacle or local area boundary, ... The search process terminates with the axis direction increasing, and the termination conditions include: a, a) An obstacle or local boundary is encountered in the axial direction; b) The number of jump points is zero; c) The target point is found. Rule 2: It is along The axis search jumps to a point, and after encountering an obstacle or local area boundary, ... The search process terminates with the axis direction increasing, and the termination conditions include: a, a) An obstacle or local boundary is encountered in the axial direction; b) The number of jump points is zero; c) The target point is found. Heuristic functions: Heuristic functions The total cost from the starting point to the target point includes the following specific steps: S231, Calculate the distance from the starting point to the current point Search path length : S232. Calculate the cost from the current point to the target point. : S233. Calculate the obstacle cost from the current point to the target point. : S234. Calculate the total cost from the starting point to the destination. : In the formula: , , , , and Representing points respectively and points The coordinates; Indicates the starting point. This indicates the cumulative search path length from the starting point to the current point. This represents the square root function. , These represent calculating the distance from the current point to the target point. axis, The obstacle rate function of the axis, , These represent the number of obstacle grids and the total number of grids, respectively. S24: Combine the paths of the various local regions into a complete path; S3. Inflection Point Path Optimization: This includes the following steps: S31: Select the inflection points in the complete path; S32: Obtain adjacent control points based on inflection points. S33: Optimize inflection points in the path using B-spline curves; S34: Obtain the optimized path, and combine it with the original path to obtain the final path.

2. The multi-objective path planning method based on a two-dimensional grid map as described in claim 1, characterized in that, In step S24, the complete path includes sub-paths of each local region, which are recombined based on the path point positions; the complete path Represented as: In the formula: Indicates the number of subpaths. Indicates the first The number of path points in each sub-path; Full path The calculation includes the following specific steps: S241. Calculate the path in the local region. : S242, Calculate the total path : In the formula: Indicates path concatenation. This refers to the MTJPS function, used to generate... and Subpaths between Used to restore waypoints to a high-resolution cost map superior.

3. The multi-objective path planning method based on a two-dimensional grid map as described in claim 1 or 2, characterized in that, In step S31, an inflection point is a point that forms a right angle with an adjacent path point. Represented as: In the formula: , indicating the number of path points; inflection point The determination includes the following specific steps: S311, Calculation Path Angle : S312, Calculation Path Angle : S313. Determine if the current point is an inflection point: In the formula: It is an inverse trigonometric function. It is an absolute value function. This is the inflection point determination function.

4. The multi-objective path planning method based on a two-dimensional grid map as described in claim 3, characterized in that, In step S32, the control point is generated from the inflection point. Represented as: In the formula: The number of control points is a variable parameter. Control Points The calculation includes the following specific steps: S321, Calculation Path Coordinates of the control point of the direction : ， S322, Calculation Path Coordinates of the control point of the direction : ， S323. Calculate the total control points: In the formula: , They represent the calculated control points. shaft and A function of axis coordinate values This indicates control point splicing. This represents the unit distance from the control point to the inflection point and is a variable parameter.

5. The multi-objective path planning method based on a two-dimensional grid map as described in claim 4, characterized in that, In step S24, the final path includes the inflection point optimized path and the original path. Represented as: In the formula: Indicates the number of path points; Final path The calculation includes the following specific steps: S241. Calculate the inflection point optimization path: S242. Calculate the total path: In the formula: This represents the optimization function for B-spline curves. This function represents the recombination of the original path and the optimized path at the inflection point.

Images