Environment modeling and path storage method for path planning of deformable two-state detection robot

Abstract

The present invention discloses an environment modeling and path storage method for path planning of a deformable two-state detection robot, including environment modeling, basic path planning and path secondary planning, path compression transmission, path decompression and the robot executing the decompressed path; in terms of environment modeling, a "quadratic grid method" environment modeling is adopted, and the basic path planning work is completed by an existing cockroach algorithm (see authorized patent ZL201710242657.2); wherein, the real robot involved is a robot whose appearance size is changed by deformation technology, and it can pass through a narrow area by deformation; the present invention realizes a path planning scheme for a robot with changeable size, realizes more possibilities for the application of the robot, the robot can pass through a narrow space in its original state, and can detect a larger range in its deformed state.

Description

Technical Field

[0001] The invention relates to robot path planning, and in particular to an environment modeling and path storage method for path planning of a deformable two-state detection robot. Background Art

[0002] Robot path planning is one of the important research directions in the field of robotics. This research direction has a wide range of applications, such as nuclear power accident rescue, industrial spraying, material transportation in hazardous areas, logistics distribution, scientific research and exploration of alien planets, etc. Robot path planning is mainly divided into global path planning and local path planning. Global path planning requires mastering all environmental information and performing path planning based on all information in the environmental map; local path planning only requires sensors to collect environmental information in real time, understand environmental map information, and then determine the location of the map and its local obstacle distribution, so as to select the optimal path from the current node to a sub-target node. From the perspective of whether the obstacle information is obtained statically or dynamically, global path planning belongs to static planning (also known as offline planning), and local path planning belongs to dynamic planning (also known as online planning).

[0003] At present, robot path planning algorithms have been widely studied, but most of these algorithms are based on fixed-form robots. Fixed-form robots often do not perform well when dealing with complex environments. For example, due to the size of the robot, it is impossible to achieve optimal path planning through narrow passages. Therefore, deformable robots are an important development direction. At present, research on path planning based on deformable robots is still relatively small and belongs to a relatively new field. Summary of the invention

[0004] The purpose of the present invention is to provide an environment modeling and path storage method for path planning of a deformable two-state detection robot, so as to adapt to more complex working environments and realize more application possibilities of mobile robots.

[0005] The technical solution of the present invention is: the environment modeling and path storage method for path planning of a deformable two-state detection robot includes environment modeling and basic path planning. The environment modeling adopts a "quadratic grid method" environment modeling, and the basic path planning work is completed by the existing cockroach algorithm (see the authorized patent ZL2017 10242657.2); wherein the real robot involved is a robot whose appearance size is changed by deformation technology, and it can pass through a narrow area by deformation; it is characterized in that the steps of the environment modeling and path storage method are as follows:

[0006] Step 1: Complete environment modeling based on the grid method

[0007] Based on the image information of the real workspace, a grid map Map with X rows and Y columns is established. The upper left corner of the grid map is used as the origin coordinate, and the density of the generated grid map is X×Y. The cell side length is designed according to the ratio of the robot's original state to the real workspace. In the initial state, the cells in the obstacle area of ​​the grid map are marked as "0", called "obstacle cells"; the cells in the feasible area are marked as "1", called "feasible cells", marking the starting point position S and the target point position D. The cells in the grid map are taken as set elements, and the grid map is described as follows:

[0008]

[0009] In the above formula, the parameter (x, y) represents the coordinates of the cell; i is the cell subscript; the parameter m indicates whether the cell is an obstacle cell; when e i m = 1, indicating a feasible cell; when e i m = 0, indicating an obstacle cell;

[0010] Step 2: Complete path planning through the cockroach algorithm (see authorized patent ZL2017 1 0242657.2)

[0011] The optimal path based on the current grid density is obtained and recorded by the linear table Path. The Path description is as follows:

[0012]

[0013] In the above formula, a i represents the cell in the planned path that the robot will walk; L represents the path length; s is the state parameter, indicating the robot state; a i ·s represents a i The s parameter of the cell; when a i ·s=1, indicating that the robot is in the original state; when a i s=2, indicating that the robot is in a transformed state; a of all elements in the initial Path i s is assigned the value 1;

[0014] Step 3: Path quadratic planning

[0015] For the optimal path Path planned in step 2, complete the secondary planning and traverse all cells a in Path in turn i (i=1, ..., L);

[0016] Step 4: Path compression transmission

[0017] The central control end compresses the path and transmits it to the robot. The compression method is as follows:

[0018] p i =((a i ·y-1)×X+a i ·x)×10+a i ·s

[0019] Compressed path Path * The data structure is a linear table, described as:

[0020] Path * ={p i |i=1,2,…,L}

[0021] Step 5: Unzip the path

[0022] After the robot obtains the compressed path, it performs decompression as follows:

[0023]

[0024] In the above formula, mod represents the modulo operation, and div represents the integer operation;

[0025] Step 6: The robot executes the unzipped path

[0026] The robot moves along the path while passing a i The s parameter completes the control operation of deformation or restoration of the original state.

[0027] Among them, step 3 path secondary planning performs the following operations:

[0028] Step 3.1: Scan cell a i (x,y) In the Map, three adjacent cells, e (x,y-1) 、e (x-1,y-1) 、e (x-1,y) ; If the above three cells are all feasible cells, then the deformation domain T i The construction is as follows:

[0029]

[0030] Step 3.2: If the deformation domain T i If it has been generated, the Path adjustment is completed according to the linear table deletion operation; judgment: if a i+1 (x,y) ∈T i If true, execute: Delete a in Path i+1 elements, and L←L-1;

[0031] Step 3.3: If the deformation domain T i If it has been generated, modify the status parameters: a i·s←2; means the robot can transform to state 2 here.

[0032] The present invention adopts the above technical solution, which has the following advantages:

[0033] (1) A robot path planning scheme with variable size is realized, which opens up more possibilities for robot application. The robot can pass through narrow spaces in its original state and can detect a wider range in its transformed state.

[0034] (2) During the path quadratic planning process, while constructing the deformation domain, some unnecessary cells in the planned path will be reduced, thus simplifying the path information.

[0035] (3) The proposed deformable two-state robot path planning method is based on the traditional grid method, which is easy for technical personnel to understand. The method is simple and easy to implement.

[0036] (4) The transmission information of the path in the present invention is obtained by compressing the robot's state information and position information into a single integer through technical means. The compression method is simple and efficient, and can specifically save 2 / 3 of the storage space, greatly shorten the path transmission time, and reduce the required memory storage space. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is the method execution process;

[0038] Figure 2 It is a grid method environment modeling;

[0039] Figure 3 is the generated raster map;

[0040] Figure 4 The path planning effect is completed based on the cockroach algorithm;

[0041] Figure 5 is the path quadratic planning effect;

[0042] Figure 6 It is the specific implementation process of path quadratic planning. DETAILED DESCRIPTION

[0043] Below in conjunction with the accompanying drawings and Examples, the technical scheme of the present invention is described in detail, but it should not be construed as limiting the technical scheme. In the following description, a large number of specific details are provided to provide a more thorough understanding of the present invention. However, for those skilled in the art, the present invention can be implemented without one or more of these details. In other examples, in order to avoid confusion with the present invention, some technical features well known in the art are not described.

[0044] like Figure 1As shown in the figure, the execution process of this path information compression transmission method is given; please refer to Figure 1 , the following is a detailed description of each step in the method:

[0045] Step S101, complete environmental modeling based on the grid method: based on the real workspace image information, establish a grid map Map with X rows and Y columns, such as Figure 2 As shown in the figure: the upper left corner of the grid map is taken as the origin coordinate, and the density of the generated grid map is X×Y; the cell side length is designed according to the ratio of the original state of the robot and the real workspace; in the initial state, the cell 100 in the obstacle area of ​​the grid map is marked as "0", called "obstacle cell"; the cell 200 in the feasible area is marked as "1" and called "feasible cell"; the starting point 300 position S and the target point 400 position D are marked; after modeling, the grid map is generated as shown in the figure: Figure 3 As shown; taking the cells in the grid map as collection elements, the grid map is described as follows:

[0046]

[0047] In the above formula, the parameter (x, y) represents the coordinates of the cell; i is the cell subscript; the parameter m indicates whether the cell is an obstacle cell; when e i m = 1, indicating a feasible cell; when e i m = 0, indicating an obstacle cell;

[0048] Step S102, using the existing cockroach algorithm to plan the path: The existing cockroach algorithm (see the authorized patent ZL2017 1 0242657.2) is used to complete the path planning and obtain the optimal path 500 based on the current grid density, such as Figure 4 As shown, it is recorded by the linear table Path, and the Path description is shown in formula (1):

[0049]

[0050] In formula (1), a i represents the cell in the planned path that the robot will walk; L represents the path length; s is the state parameter, indicating the robot state; a i ·s represents a i The s-parameters of the cell; e.g. Figure 5 As shown: When a i ·s=1, indicating that the robot is in the original state 600; when a i ·s=2, indicating that the robot is in the transformed state 700; a of all elements in the initial Path i s is assigned the value 1;

[0051] Step S103, path secondary planning: for the optimal path Path planned in step 2, complete the secondary planning and traverse all cells a in Path in turn i (i=1, ..., L), complete the robot deformation parameter setting, and generate the secondary planning path 800;

[0052] Step S104, path compression transmission: The central control end compresses the path and transmits it to the robot. The compression method is shown in formula (2):

[0053] p i =((a i ·y-1)×X+a i ·x)×10+a i ·s (2)

[0054] Compressed path Path * The data structure is a linear table, as shown in formula (3):

[0055] Path * ={p i | i=1,2,…,L} (3)

[0056] Step S105, path decompression: After the robot obtains the compressed path, it performs decompression, and the method is shown in formula (4):

[0057]

[0058] In the above formula, mod represents the modulo operation, and div represents the integer operation;

[0059] Step S106: The robot executes the decompressed path: while the robot moves along the path, it passes through a i The s parameter completes the control operation of deformation or restoration of the original state.

[0060] Combine the following Figure 6 The flowchart of step S103 is described in detail; the path secondary planning performs the following specific operations:

[0061] Step S201: set loop variable i←0;

[0062] Step S202: Execute: i←i+1;

[0063] Step S203: Construct deformation domain T i ; Scan cell a i (x,y) In the Map, three adjacent cells, e (x,y-1) 、e (x-1,y-1) 、e (x-1,y); If the above three cells are all feasible cells, then the deformation domain T i Existence, deformation domain T i The construction is as follows:

[0064]

[0065] Step S204: Determine if the deformation domain T i If the construction is successful, execute S205; otherwise: execute S207;

[0066] Step S205: Complete the path adjustment according to the linear table deletion operation; judge: if a i+1 (x,y) ∈T i If true, execute: Delete a in Path i+1 elements, and L←L-1;

[0067] Step S206: Modify state parameters: a i s←2; means the robot can transform to state 2 here;

[0068] Step S207: If i<L is true, then execute: jump to step S202; otherwise execute S208;

[0069] Step S208: Output the secondary planning path.

Claims

1. The environmental modeling and path storage method for the path planning of a deformable two-state detection robot includes environmental modeling and basic path planning. The environmental modeling adopts a "quadratic grid method" environmental modeling, and the basic path planning work is completed by the existing cockroach algorithm; among them, The real robot involved is a robot that changes its appearance size through deformation technology, and can pass through narrow areas through deformation; its characteristics are that the steps of the environment modeling and path storage method are as follows: Step 1: Complete environment modeling based on the grid method Based on the real workspace image information, a grid map Map with X rows and Y columns is established. The upper left corner of the grid map is used as the origin coordinate, and the density of the generated grid map is X×Y. The cell side length is designed according to the ratio of the robot's original state to the real workspace. In the initial state, the cells in the obstacle area of ​​the grid map are marked as "0", called "obstacle cells"; the cells in the feasible area are marked as "1", called "feasible cells", marking the starting point position S and the target point position D. The cells in the grid map are taken as set elements, and the grid map description is as follows: In the above formula, the parameter (x, y) represents the coordinates of the cell; i is the cell subscript; the parameter m indicates whether the cell is an obstacle cell; when e i m = 1, indicating a feasible cell; when e i m = 0, indicating an obstacle cell; Step 2: Complete path planning through the cockroach algorithm The optimal path based on the current grid density is obtained and recorded by the linear table Path. The Path description is as follows: In the above formula, a i represents the cell in the planned path that the robot will walk; L represents the path length; s is the state parameter, indicating the robot state; a i ·s represents a i The s parameter of the cell; when a i ·s=1, indicating that the robot is in the original state; when a i s=2, indicating that the robot is in a transformed state; the a of all elements in the initial Path i s is assigned the value 1; Step 3: Path quadratic planning For the optimal path Path planned in step 2, complete the secondary planning and traverse all cells a in Path in turn i , where i = 1, ..., L; The specific steps are as follows: Step S101, complete environmental modeling based on the grid method: based on the real workspace image information, establish a grid map Map with X rows and Y columns, the upper left corner of the grid map is used as the origin coordinate, and the density of the generated grid map is X×Y; the cell side length is designed according to the ratio of the original state of the robot and the real workspace; in the initial state, the cell 100 in the obstacle area of ​​the grid map is marked as "0", called "obstacle cell"; the cell 200 in the feasible area is marked as "1" called "feasible cell"; mark the starting point 300 position S and the target point 400 position D; generate a grid map after modeling; take the cells in the grid map as set elements, and the grid map description is as follows: In the above formula, the parameter (x, y) represents the coordinates of the cell; i is the cell subscript; the parameter m indicates whether the cell is an obstacle cell; when e i m = 1, indicating a feasible cell; when e i m = 0, indicating an obstacle cell; Step S102, using the existing cockroach algorithm to plan the path: The existing cockroach algorithm is used to complete the path planning, and the optimal path 500 based on the current grid density is obtained, which is recorded by the linear table Path. The Path description is shown in formula (1): In formula (1), a i represents the cell in the planned path that the robot will walk; L represents the path length; s is the state parameter, indicating the robot state; a i ·s represents a i The s parameter of the cell; when a i ·s=1, indicating that the robot is in the original state 600; when a i ·s=2, indicating that the robot is in the transformed state 700; a of all elements in the initial Path i s is assigned the value 1; Step S103, path secondary planning: for the optimal path Path planned in step 2, complete the secondary planning and traverse all cells a in Path in turn i , where i = 1, ..., L, complete the robot deformation parameter setting and generate the secondary planning path 800; Step S104, path compression transmission: The central control end compresses the path and transmits it to the robot. The compression method is shown in formula (2): p i =((a i ·y-1)×X+a i ·x)×10+a i ·s (2) Compressed path Path * The data structure is a linear table, as shown in formula (3): Path * ={p i | i=1,2,…,L} (3) Step S105, path decompression: After the robot obtains the compressed path, it performs decompression, and the method is shown in formula (4): In the above formula, mod represents the modulo operation, and div represents the integer operation; Step S106: The robot executes the decompressed path: while the robot moves along the path, it passes through a i The s parameter completes the control operation of deformation or restoration of the original state; The path secondary planning in step S103 performs the following specific operations: Step S201: set loop variable i←0; Step S202: Execute: i←i+1; Step S203: Construct deformation domain T i ; Scan cells In the Map, three adjacent cells, e (x,y-1) 、e (x -1,y-1) 、e (x-1,y) ; If the above three cells are all feasible cells, then the deformation domain T i Existence, deformation domain T i The construction is as follows: Step S204: Determine if the deformation domain T i If the construction is successful, execute S205; otherwise: execute S207; Step S205: Complete the path adjustment according to the linear table deletion operation; judge: if a i+1 (x,y) ∈T i If true, execute: Delete a in Path i+1 elements, and L←L-1; Step S206: Modify state parameters: a i s←2; means the robot can transform to state 2 here; Step S207: If i<L is true, then execute: jump to step S202; otherwise execute S208; Step S208: output the secondary planning path; Step 4: Path compression transmission The central control end compresses the path and transmits it to the robot. The compression method is as follows: p i =((a i ·y-1)×X+a i ·x)×10+a i ·s Compressed path Path * The data structure is a linear table, described as: Path * ={p i |i=1,2,…,L} Step 5: Unzip the path After the robot obtains the compressed path, it performs decompression as follows: In the above formula, mod represents the modulo operation, and div represents the integer operation; Step 6: The robot executes the unzipped path The robot moves along the path while passing a i The s parameter completes the control operation of deformation or restoration of the original state.

2. The method for environment modeling and path storage for path planning of a deformable two-state detection robot according to claim 1, characterized in that: Step 3: Path quadratic planning performs the following operations: Step 3.1: Scan cells In the Map, three adjacent cells, e (x,y-1) 、e (x-1,y-1) 、e (x-1,y) ; If the above three cells are all feasible cells, then the deformation domain T i The construction is as follows: Step 3.2: If the deformation domain T i If it has been generated, the Path adjustment is completed according to the linear table deletion operation; Judgment: If a i+1 (x,y) ∈T i If true, execute: Delete a in Path i+1 elements, and L←L-1; Step 3.3: If the deformation domain T i If it has been generated, modify the status parameters: a i ·s←2; means the robot can transform to state 2 here.

Images