Efficient coverage path planning algorithm

Abstract

The invention provides an efficient coverage path planning algorithm, which comprises the following steps of: decomposing a plurality of convex polygon small-area regions by using a traditional cattle culture cell decomposition method, establishing a communication relationship between the regions, and counting the sweeping length of each region; selecting an end point area, judging whether an area adjacent to the end point area meets a unique affiliation condition or not, if yes, adding the unique affiliation area to the current area, and splicing the current affiliation area with the track of the current area; if not, an optimal traversal area is found through weighted judgment, and the optimal traversal area is spliced with the track of the current area; and carrying out area traversal from an end point area to a starting point area, connecting the spliced tracks in series, and finally generating a section of complete coverage cleaning path. According to the method, the condition weight of the current surrounding small area of the robot can be well improved, so that the traversal sequence of the cleaning area is more reasonable, and the problem of coverage area dispersion is avoided.

Description

Technical Field

[0001] The present invention relates to the field of robot path planning, and in particular to an efficient coverage path planning algorithm. Background Art

[0002] Currently, almost all laser robot vacuums on the market use a ox-plowed path-covering method to generate their cleaning paths. This ox-plowed cell decomposition method breaks down the cleaning area into several small convex polygonal areas. Within each convex polygonal area, an ox-plowed path is generated based on the cleaning width. The starting and ending points of each small area are then connected using navigation trajectories to ultimately generate a complete, covered cleaning path. Existing technologies fail to determine the weights of the small areas, resulting in an illogical order in which the robot traverses the cleaning area, impacting cleaning efficiency. Summary of the Invention

[0003] In order to solve the problems of the prior art, the present invention provides an efficient coverage path planning algorithm, which can greatly improve the situation weight of the small area around the robot, make the cleaning area traversal order more reasonable, and avoid the problem of scattered coverage areas.

[0004] The present invention provides an efficient coverage path planning algorithm, comprising the following steps:

[0005] 1) Obtain the cleaned map and preprocess the map;

[0006] 2) Use the traditional ox-plow cell decomposition method to decompose multiple convex polygonal small areas and mark them as the areas to be visited;

[0007] 3) Confirm whether the area to be visited is empty. If not, proceed to step 4); if it is empty, complete the path planning, connect the spliced ​​trajectories in series, and finally generate a complete coverage cleaning path;

[0008] 4) Establish a connection between each area and count the cleaning length of each area; areas with only one connection are considered as attached areas of the current area, i.e., broken area scenarios, and are marked as the only attached area;

[0009] 5) Select the destination area and determine whether the adjacent areas of the destination area meet the unique attachment condition. If so, add the unique attached area to the current area and splice the trajectory of the current attached area with that of the current area. If not, find the optimal traversal area through weighted judgment and splice the optimal traversal area with the trajectory of the current area.

[0010] 6) The current area is extracted from the area to be visited, and the area traversal is performed from the end area to the starting area, and then the process returns to step 3).

[0011] As a further improvement, the preprocessing process in step 1) is to smooth the map within the cleaning area to eliminate burrs in the map.

[0012] As a further improvement, the weighted judgment process in step 5) is specifically as follows: searching for up to three nearest candidate areas, marking the distance to the area as D = {d1, d2, d3}, and obtaining the area lengths L = {l1, l2, l3} of the three candidate areas, and selecting the optimal target area from the candidate areas through the weighted judgment formula.

[0013] The beneficial effect of the present invention is that it can well improve the situation weight of the small area currently around the robot, make the traversal order of the cleaning area more reasonable, and avoid the problem of scattered coverage area. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0015] Figure 1 It is the flow chart of the algorithm of the present invention;

[0016] Figure 2 Schematic diagram of the decomposition of cattle tiller cells;

[0017] Figure 3 This is a schematic diagram of the order of traversal of the cattle-ploughed area;

[0018] Figure 4 Schematic diagram of the clearing path for complete ox-plowing. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0020] The present invention will be further described below by a specific embodiment. The map used in the specific embodiment is as follows: Figure 1 shown.

[0021] Step 1: Smooth the map within the cleaning area through map preprocessing. This eliminates burrs in the map and avoids excessive small areas when performing ox-plow cell decomposition.

[0022] Step 2: Use the traditional ox-farming cell decomposition method to decompose multiple convex polygonal small areas, such as Figure 2 shown.

[0023] Depend on Figure 3 As you can see, five regions appear within the box. If each region is connected using the closest navigation distance method (why regions are generally connected using the closest navigation distance method? Because the TSP method is computationally intensive, while the closest navigation distance method is simple and stable), then the coverage area may be fragmented.

[0024] For example, the traversal order could be 1, 2, 3, 5, 4, or 1, 5, 2, 3, 4. Either of these is undesirable, because area 5 may be very large. The robot may run far away to clean area 5 and then return to clean a small area like areas 3 and 4.

[0025] Step 3: Establish a connectivity relationship (Connectives) between each area. For example, the connectivity relationship of 1 is 2 and 5. The connectivity relationship of 2 is 1, 3, and 4. The connectivity relationship of 3 is 2. Count the cleaning length of each area (Lengths), where the length unit is m.

[0026] Connectives=Connective[N]

[0027] Connective[1]={2 5}

[0028] Connective[3]={1 3 4}

[0029] Connective[3]={2}

[0030] Lengths=Length[N]

[0031] Length[1]=20.6m

[0032] Length[3]=1.5m;

[0033] Step 4: When selecting regions for traversal, treat regions with only one connection as dependent regions of the current region. This is because these regions can only be entered from the current region, creating a disconnected region scenario. These regions are marked as unique dependent regions. For example, if 3 and 4 are the only dependent regions of 2, change the length attribute of these regions to Length[2] = Length[3] + Length[4].

[0034] Step 5: In addition to the unique attached region feature, for regions with regional connectivity greater than 2, a weighted judgment method is used. A breadth-first search is used to identify up to three nearest candidate regions, and the distance to each region is marked as D = {d1, d2, d3}. The region lengths of the three candidate regions are then calculated as L = {l1, l2, l3}. The optimal target region is selected from the candidate regions using the weighted judgment formula: cell = min(d + k * l), where k is the region length weight.

[0035] Step 6: Starting from the end area, judge the areas one by one and connect the trajectories to finally generate a complete coverage cleaning path, such as Figure 4 shown.

[0036] This method can effectively increase the weight of the small area around the robot, making the cleaning area traversal order more reasonable and avoiding the problem of scattered coverage areas. In this way, even if the end point of area 1 is in the lower left corner, the traversal order will still be 1, 2, 3, 4, 5.

[0037] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, the above is only a preferred embodiment of the present invention. Since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited to this. Any technical personnel familiar with this technical field is within the technical scope disclosed by the present invention. For ordinary technical personnel in this technical field, changes or replacements that can be easily thought of should be covered within the protection scope of the present invention without departing from the principle of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. An efficient coverage path planning algorithm, characterized by The following steps are involved: 1) Obtain a cleanup map within the task area and pre-process the map to eliminate burrs on the map and prevent the appearance of too many small areas when performing subsequent ox-plow cell decomposition; 2) Use the traditional ox-plow cell decomposition method to decompose multiple convex polygonal small areas and mark them as the areas to be visited; 3) Confirm whether the area to be visited is empty. If not, proceed to step 4); if it is empty, complete the path planning, connect the spliced ​​trajectories in series, and finally generate a complete coverage cleaning path; 4) Establish a connection between each area and count the cleaning length of each area; areas with only one connection are considered as attached areas of the current area, i.e., broken area scenarios, and are marked as the only attached area; Based on the robot's current position, a breadth-first search is performed to find the first nearest ox-plowed area as the starting coverage area, and the end point of this area is used as the starting point for searching the next coverage area. 5) Select the next coverage area and determine whether the adjacent area of ​​the next coverage area meets the unique attachment condition. If so, add the unique attached area to the current area and splice the trajectory of the current attached area with that of the current area; If it is not satisfied, the optimal traversal area is found through weighted judgment, and the trajectory of the optimal traversal area is spliced ​​with the trajectory of the current area; 6) Remove the current area from the area to be visited and return to step 3).

2. The efficient coverage path planning algorithm according to claim 1, characterized in that: The pre-processing process in step 1) is to smooth the map in the cleaning area and eliminate burrs in the map.

3. The efficient coverage path planning algorithm according to claim 1, characterized in that: Step 5) The weighted judgment process is specifically as follows: search for up to 3 nearest candidate areas, mark the distance to the area as D = {d1, d2, d3}, and obtain the area length L = {l1, l2, l3} of the 3 candidate areas, and select the optimal target area from the candidate areas through the weighted judgment formula.

Images