Mobile robot full-coverage path planning algorithm based on energy consumption

Abstract

The invention relates to a mobile robot full-coverage path planning algorithm based on energy consumption, and the algorithm comprises the steps: calculating the energy consumption of a neighbor available grid from a starting point, and determining a next grid according to Euclidean distance constraint, steering angle constraint, surrounding available grid constraint, optimal motion direction constraint and a grid selection sequence; if the current grids are all covered or obstacles exist at the moment, but uncovered grids still exist in the whole area, the dead zone is escaped through a backtracking-based dead zone escaping algorithm, the covered grids are traversed in an inverted order, the first uncovered grid around the covered grids is found, and the grid is the dead zone escaping grid. And after the dead zone is escaped, the previous full coverage is continued until all traversal of the grids in the whole specified area is completed. The method can ensure that the planned full-coverage path is shorter and the repetition rate is lower.

Description

Technical Field

[0001] The present invention relates to the field of path planning for cleaning robots, and particularly to a full-coverage path planning algorithm for mobile robots based on energy consumption. Technical Background

[0002] With the development of mobile robot technology, the application fields of cleaning robots are more extensive, such as cleaning in parking lots, cleaning urban roads, etc., with broad application prospects. Key technologies involved in cleaning robots, such as map construction, positioning and navigation, and path planning, are all hotspots in the research of mobile robots. The main purpose of full-coverage path planning is to minimize the path repetition rate and total length of the algorithm and reduce the number of times of getting stuck in dead zones on the basis of ensuring a high coverage rate of the specified area. Full-coverage path planning is the most direct manifestation of the intelligent level of cleaning robots, directly affecting the working efficiency of cleaning robots and being one of the core issues of cleaning robots. Summary of the Invention

[0003] Aiming at the problems of low coverage rate, high repetition rate, and long total planned path of existing full-coverage path planning algorithms in complex environments, a full-coverage path planning algorithm for cleaning robots based on energy consumption is proposed. The planned algorithm reduces the number of times of getting stuck in dead zones, the path length, and the path repetition rate while meeting the requirements of the coverage rate of the cleaning area.

[0004] The present invention adopts the following technical solutions, and the steps include:

[0005] Step 1: Obtain the environment where the target object is located. The environmental information includes the starting point, the target area, obstacles, passable areas, and a container path for storing the planned path.

[0006] Step 2: Calculate the available neighbor grids around the current grid

[0007] By exploring the neighbor grids around the current grid and screening out all available neighbor grids k according to whether the grid is covered or is an obstacle. If the number of current available neighbor grids is 0, then calculate the number of uncovered grids in the entire cleaning area If the number of is 0, it means that the whole area has been covered and the cleaning task is completed; if the number of current available neighbor grids is 0, the number of is not 0, it means that the robot is stuck in a dead zone at this time, that is, the neighbor area of the current grid has been explored but the whole area has not been cleaned yet. At this time, enter Step 3; if the number of current available neighbor grids is not 0, then enter Step 4.

[0008] Step 3: Plan an algorithm to escape from the dead zone

[0009] When the robot gets stuck in a dead zone, traverse the grids in the list in reverse order according to the currently covered path list "path", select the neighbor grids of the grids in the backtracking list according to the neighbor grid selection rules established in the planning algorithm, until the first uncovered neighbor grid is selected. This grid is the optimal dead zone escape point. Then, use the A* algorithm to plan the path from the current position to the dead zone escape grid. After planning the path, the robot can escape from the dead zone and then continue to execute step two.

[0010] Step Four: Calculate the energy of the available grids among the eight neighborhood grids around the current grid in a certain exploration order and through the energy consumption formula, and select the grid with the least energy consumption. This grid is the next planned grid.

[0011] E(x,y) = d t (x,y) + d r (x,y) + 0.3*d n (x,y) + d q (x,y)(1)

[0012] Among them, d t (x,y) represents the Euclidean distance constraint of the neighborhood grid, d r (x,y) represents the direction confidence constraint of the neighborhood grid, d n (x,y) represents the neighborhood coverage grid constraint of the neighborhood grid, d q (x,y) represents the optimal motion direction constraint of the neighborhood grid.

[0013] Furthermore, according to the above definitions, the exploration order is down, left, right, up, lower left, lower right, upper left, upper right;

[0014] Furthermore, according to the above definitions, the specific details of each constraint are:

[0015]

[0016] Among them, l(x l ,y l ) represents the coordinates of an available neighbor grid of the current grid, c(x c ,y c ) represents the coordinates of the grid at the current position.

[0017]

[0018] Among them, l θ is the direction from the current grid to the neighbor grid, c θ is the current direction of the robot. For the convenience of calculation, the included angle between the two can be obtained by the following formula, where the grid cp(x cp ,y cp)The previous grid representing the current grid, the direction vector from the previous grid to the current grid represents the current direction of the robot. The direction vector from the current grid to the next grid represents the orientation of the robot's next step, and the included angle between the two is the turning angle of the robot.

[0019] l θ -c θ = arccos<(x l -x c ,y l -y c ),(x c -x cp ,y c -y cp )> (4)

[0020]

[0021] Where k represents the grids that have been covered around the neighbor grids of the current grid, path represents the set of grids that have been covered, representing all available neighbor grids.

[0022]

[0023] Calculate the number of continuous regions increased due to obstacles in all rows and all columns of the grid area, and the direction represented by the smaller value of the two is the specified overall operation direction.

[0024] Step Five: Compare the energies of all neighbor grids of the current grid, and the one with the minimum energy consumption is the next grid, then go back to Step Two for the next full-coverage exploration.

[0025] Step Six: Repeat the above steps until the full-coverage path planning for the given area is completed.

[0026] The advantages of the present invention are as follows: The present invention provides a full-coverage path planning algorithm based on energy consumption, which expands and optimizes the existing full-coverage path planning technology. It not only considers the coverage rate of the given area, but also takes into account the path repetition rate and path length; the full-coverage path planning method proposed by the present invention can automatically plan a path that simultaneously meets high coverage and low repetition rate in a grid map environment, and the planned path is more regular, providing the possibility for the robot to better perform the cleaning task. A path that makes the robot move more coherently and can simplify the smoothing operation is obtained. Description of the Drawings

[0027] To more clearly illustrate the technical solutions of the embodiments of the patent application of the present application, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the patent application of the present application. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0028] Figure 1 It is a flowchart of a full-coverage path planning algorithm based on energy consumption;

[0029] Figure 2 It is a process diagram of the specific planning of the full-coverage path planning algorithm based on energy consumption. Specific embodiments

[0030] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0031] As Figure 2 (a), the black area in the figure is an obstacle, the red is the starting point, and the white area is the area to be covered. The purpose of the present invention is to plan a route with a shorter path length and lower repetition rate that covers all white areas.

[0032] As Figure 2 (b) shows the path. Starting from the starting point, first explore the neighboring grids around the current grid, calculate the next point according to the above formula and its meaning, and loop in turn until the overall planning is completed or a dead zone is encountered.

[0033] As Figure 2 (b) shows that at this time, the planned route enters a dead zone. According to the dead zone escape algorithm, traverse the grids path that have been walked in reverse order, and find the first uncovered grid around these grids. This grid is the grid for escaping the dead zone.

[0034] As Figure 2 (c) shows that after the full-coverage path planning algorithm escapes from the dead zone, it continues to run. When it enters the dead zone, it executes the dead zone escape algorithm, and after escaping from the dead zone, it continues to execute the full-coverage planning algorithm until the full-coverage task for the given area is completed.

Claims

1. A mobile robot full-coverage path planning algorithm based on energy consumption, characterized in that, It includes the following steps: Step 1: Obtain the environment where the target object is located. The environmental information includes the starting point, the target area, obstacles, the passable area, and the container path for storing the planned path. Step 2: Calculate the available neighbor grids around the current grid By exploring the neighboring grids around the current grid and filtering out all available neighboring grids k based on whether the grid is covered or is an obstacle. If the number of currently available neighboring grids is 0, then calculate the number of uncovered grids in the entire cleaning area. If the number is 0, it means the entire area has been covered and the cleaning task is completed; if the number of currently available neighboring grids is 0, the number is not 0, it means the robot is in a dead end at this time, that is, the neighboring area of the current grid has been explored but the entire area has not been cleaned yet. At this time, go to step three. If the number of currently available neighboring grids is not 0, then go to step four; Step 3: Plan the dead zone escape algorithm When the robot falls into the dead zone, based on the current covered path list path, traverse the grids in the list in reverse order, and select the neighbor grids of the grids in the backtracking list according to the neighbor grid selection rules established in the planning algorithm until the first uncovered neighbor grid is selected. This grid is the optimal dead zone escape point. Then, plan the path from the current position to the dead zone escape grid through the A* algorithm. After planning the path, the dead zone can be escaped, and then Step 2 is continued. Step 4: Calculate the energy of the available grids in the eight neighborhood grids around the current grid in a certain exploration order and through the energy consumption formula, and select the grid with the least energy consumption. This grid is the next planned grid. E(x, y) = d t (x, y) + d r (x, y) + 0.3*d n (x, y) + d q (x, y) (1) Step 5: Compare the energy of all the neighbor grids of the current grid. The one with the least energy consumption is the next grid, and then go back to Step 2 for the next full-coverage exploration. Step 6: Repeat the above steps until the full-coverage path planning of the given area is completed.

2. The full-coverage path planning algorithm according to claim 1, wherein Its exploration order is: Down, left, right, up, lower left, lower right, upper left, upper right.

3. The full-coverage path planning algorithm according to claim 1, characterized in that, The descriptions of each part of the energy calculation function E(x, y) are as follows: where l(x l ,y l ) represents an available neighbor grid coordinate of the current grid, c(x c ,y c ) represents the grid coordinates of the current position. l θ -c θ = arccos < (x l -x c , y l -y c ), (x c -x cp , y c -y cp ) > (4) where l θ is the direction from the current grid to the neighboring grid, and c θ is the current direction of the robot. For convenience of calculation, the included angle between the two can be obtained using the following formula. Among them, the grid cp(x cp , y cp ) represents the previous grid of the current grid. The direction vector from the previous grid to the current grid represents the current direction of the robot. The direction vector from the current grid to the next grid represents the orientation of the robot's next step. The included angle between the two is the turning angle of the robot. Where k represents the grids that have been covered around the neighbor grids around the current grid, path represents the set of grids that have been covered, and represents all available neighbor grids. First, calculate the number of consecutive areas that are more due to obstacles for all rows and all columns in the grid area. The direction represented by the smaller value of the two is the specified overall running direction (x z , y z ).

Images