A grinder full-area coverage path planning method
By decomposing a secure grid map and planning a room adjacency graph, smooth path connections are generated, solving the problems of incomplete path coverage and unstable equipment operation in complex environments in existing technologies, and achieving efficient and secure path coverage.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SUZHOU GENGXUN INTELLIGENT TECHNOLOGY CO LTD
- Filing Date
- 2026-03-20
- Publication Date
- 2026-06-05
AI Technical Summary
Existing grinding machine path planning methods are difficult to effectively cover all passable areas in complex environments, resulting in repeated grinding or omissions, low path connection efficiency, and insufficient consideration of equipment kinematic constraints and collision avoidance, which affects equipment operating efficiency and safety.
By acquiring a safe grid map, decomposing it into connected component rooms, constructing a room adjacency graph, generating smooth sweep segments, using the A* algorithm to connect paths, and performing path straightening and collision avoidance checks to ensure that the path covers all passable areas and satisfies kinematic constraints.
It achieves efficient coverage of all passable areas in complex environments, improves the structuring and computational efficiency of path planning, reduces the number of turns, and enhances the stability and safety of equipment operation.
Smart Images
Figure CN122149476A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of mobile robot path planning technology, and in particular to a method for full-area coverage path planning for a grinding machine. Background Technology
[0002] With the development of automated equipment and intelligent robot technology, floor grinding equipment is gradually moving towards automation and intelligence. Automatic grinding machines are widely used in industrial plant floor treatment, warehouse and logistics site maintenance, and large-scale building floor construction. In practical applications, grinding machines need to complete grinding operations on large areas of the floor in environments containing obstacles such as walls, equipment, and columns. Therefore, how to plan a path that covers the entire passable area and has high movement efficiency becomes a crucial issue affecting the equipment's operating efficiency.
[0003] In existing technologies, the path planning methods for grinding equipment mostly employ simple reciprocating scanning or manually set paths, such as parallel reciprocating movement in a fixed direction. However, when there are multiple obstacles or complex spatial structures within the work area, the above methods are prone to the following problems: First, it is difficult to effectively decompose complex and irregular passable areas, leading to repeated grinding or omissions in some areas; second, the path connection efficiency is low when switching between different areas, easily resulting in a large amount of ineffective movement; third, the path contains many broken lines and sharp turns, which is not conducive to the smooth operation of the equipment and affects the grinding quality and work efficiency.
[0004] Furthermore, while some path planning methods can achieve area coverage, they do not fully consider the kinematic constraints of the equipment during path generation, such as minimum turning radius or path smoothness. This can lead to problems such as difficult turns or frequent decelerations during actual execution of the planned path. Additionally, the lack of an effective collision avoidance mechanism during path optimization may cause the optimized path to approach or enter obstacle areas, thereby affecting the safety of equipment operation.
[0005] Therefore, how to achieve reasonable decomposition of passable areas in complex environments containing obstacles, and generate efficient paths that can cover all passable areas while ensuring path passability and smooth movement, has become a technical problem that urgently needs to be solved in the field of grinding machine path planning technology. Summary of the Invention
[0006] The purpose of this invention is to address the problems existing in the background art by proposing a full-area coverage path planning method for a grinding machine.
[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for full-area coverage path planning of a grinding machine, comprising the following steps: Step S1: Map acquisition and preprocessing. Acquire a grid map containing passable areas and obstacles, and inflate the obstacles in the grid map according to the equivalent radius of the grinder to obtain a safety grid map; Step S2: Room decomposition and representation. Connected components are marked for the passable areas in the safety grid map. Each connected component is treated as a room, and a bounding rectangle is determined for each room as its geometric representation. Step S3: Room access order planning. Construct a room adjacency graph based on the positional relationship between the bounding rectangles of each room, and traverse the room adjacency graph starting from the room where the grinding machine originates to determine the room access order; Step S4: Generating sweeping paths inside the room, processing each room sequentially according to the access order, determining the sweeping direction based on the aspect ratio of the room's circumscribed rectangle, and generating parallel sweeping segments covering the passable area of the room according to a preset row spacing related to the grinding width of the grinder; Step S5: Path connection, using the A* algorithm to calculate the connection paths between adjacent sweep segments and between different rooms on the security grid map; Step S6: Path straightening optimization. The straight connectivity of the connecting path is checked, and the path is straightened and simplified by finding the farthest point in the path sequence that is directly connected to the current path point. The straight connectivity is verified for drivability using the Bresenham straight line algorithm. Step S7: Path geometric smoothing, compressing collinear points in the path, thinning out small-angle inflection points, and inserting circular arc transition segments at the retained inflection points according to the minimum turning radius of the grinder to generate an executable path that satisfies kinematic constraints; Step S8: Collision avoidance verification. Perform drivability detection on the smoothed path to ensure that all points on the path are within the drivable area of the safety grid map, and output a full-area coverage path for the grinder that can cover the drivable area in the safety grid map.
[0008] Furthermore, the circumscribed rectangle is an axis-aligned rectangle. Its world coordinates are obtained by statistically analyzing the grid row and column ranges of the corresponding connected components and combining the grid map resolution with the origin coordinate transformation. The resolution of the grid map is set according to the size of the grinder and the positioning accuracy, so that the side length of a single grid is not greater than the minimum safe gap of the grinder.
[0009] Furthermore, when constructing the room adjacency graph, the minimum gap between the bounding rectangles of any two rooms in the X-axis and Y-axis directions is calculated. When the minimum gap in both directions is less than or equal to the preset adjacency threshold, the two rooms are determined to be adjacent.
[0010] Furthermore, the room adjacency graph is traversed using a depth-first traversal method. When the room adjacency graph is a disconnected graph, untraversed rooms are added to the end of the access order according to the minimum gap between their bounding rectangles and the bounding rectangles of the traversed room set.
[0011] Furthermore, when the width of the circumscribed rectangle of the room is greater than or equal to its height, a sweep segment extending horizontally is generated; when the width of the circumscribed rectangle is less than its height, a sweep segment extending vertically is generated.
[0012] Furthermore, after generating the sweep segment, the passability of the sweep segment is checked. When the line connecting the first and last points of the sweep segment is passable, only the first and last points are retained as path points. When there are impassable grids, grids are sampled along the sweep segment direction and the center point of the passable grid is retained as a path point.
[0013] Furthermore, when using the A* algorithm for pathfinding, priority is given to pathfinding within the combined bounding rectangle of the rooms where the starting and ending points are located. If pathfinding fails, the range restriction is removed and pathfinding is performed on the entire safety grid map.
[0014] Furthermore, when a pathfinding for a connection fails, the corresponding target sweep segment is added back to the tail of the sweep segment connection queue, and the connection is attempted again after subsequent path planning is completed.
[0015] Furthermore, inflection point thinning is determined by calculating the angle between two adjacent path segments and the vertical distance from the path point to the line connecting the preceding and following path segments. When the angle is less than a preset angle threshold and the vertical distance is less than a preset distance threshold, the path point is deleted.
[0016] Furthermore, the collision avoidance check converts the world coordinates of the path points into raster coordinates based on the raster map resolution and the origin coordinates, and checks whether the corresponding raster is passable. When there are impassable points, the corresponding path segment is deleted and the path connection calculation in step S5 is re-executed.
[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: By marking the accessible areas in the safety grid map with connected components (each connected component is a room) and using bounding rectangles to geometrically represent each room, complex work environments can be automatically decomposed into multiple clearly structured work areas, thereby facilitating subsequent path planning and improving the structuring and computational efficiency of path planning. By constructing a room adjacency graph and determining the room access order, the grinding machine can perform operations in a reasonable order of areas, reducing cross-area movement distance and invalid travel paths, thereby improving overall work efficiency; The sweep direction is adaptively determined based on the aspect ratio of the room's circumscribed rectangle, and parallel sweep segments are generated to ensure that the sweep direction is consistent with the longer direction of the room's circumscribed rectangle. This effectively reduces the number of turns in the path and improves area coverage efficiency. By employing the A* algorithm to connect adjacent sweep segments and different rooms, and performing pathfinding while prioritizing a limited search range, path reachability can be guaranteed while reducing the computational load of pathfinding and improving path connection efficiency. By performing straight-line connectivity detection on the connection path and straightening the path, redundant intermediate path points can be reduced, making the generated path simpler, thereby reducing equipment control complexity and improving operating efficiency; By compressing collinear points in the path and thinning out small-angle inflection points, and by inserting arc transition segments at key inflection points based on the minimum turning radius of the grinding machine, the generated path is made smoother, better meeting the kinematic constraints of the grinding machine and improving the actual operational stability of the equipment. By performing collision avoidance checks on the final generated path, it is ensured that all points on the path are within the passable area of the safety grid map, thereby effectively avoiding collisions between the path and obstacles and improving the safety of equipment operation. Attached Figure Description
[0018] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation
[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0020] In this embodiment, the grinding machine full-area coverage path planning method can be run in the control system of a ground grinding equipment with autonomous mobility. The system can obtain the working environment information through the lidar SLAM mapping module or the existing CAD environment map, and generate a working path that can cover all passable areas on this basis.
[0021] like Figure 1 As shown, the path planning method described in this embodiment mainly includes the following steps: map preprocessing, room area decomposition, room access order planning, internal room coverage path generation, path connection, path straightening optimization, path geometric smoothing, and anti-collision verification.
[0022] Step S1: Map Acquisition and Preprocessing First, obtain a raster map of the environment to be operated on. The raster map consists of several equal-sized raster cells, each representing a small area in the environment and marked as a passable area or an obstacle area.
[0023] Because the grinder has certain physical dimensions, if path planning is done directly according to the original obstacle boundaries, the equipment may get too close to the obstacles during operation, creating a collision risk. Therefore, it is necessary to expand the obstacles based on the grinder's equivalent safety radius to generate a safety grid map.
[0024] In this embodiment, the need for dilation marking can be determined by calculating the Euclidean distance between the center point of the grid and the center point of the obstacle grid. For example, the calculation can be performed as follows: in: This indicates the distance between the center point of the current grid cell to be detected and the center point of the obstacle grid cell; , This indicates the coordinates of the current grid center point in the map coordinate system; , Represents the coordinates of the center point of the obstacle grid.
[0025] When the calculated distance satisfies: If the grid is blocked, then the grid is marked as an impassable area.
[0026] in: This indicates the equivalent safety radius of the grinding mill; This parameter is usually determined by the radius of the grinding machine body and the safety margin.
[0027] The above method can create a safety buffer zone around the original obstacle, thus ensuring that the grinding machine has a sufficient safe distance when planning its path.
[0028] The resolution of the grid map is set according to the size of the grinder and the positioning accuracy, so that the side length of a single grid cell is no greater than the minimum safe gap of the grinder, thus ensuring the rationality of grid division and the safety of path planning.
[0029] Step S2: Room Decomposition and Area Representation After obtaining the safety grid map, the passable areas in the map are marked as connected components. Each connected component is treated as an independent room area.
[0030] To facilitate subsequent path planning calculations, this embodiment uses an axis-aligned circumscribed rectangle as the geometric representation of the room.
[0031] Let G be the set of grid cells contained in a room, where the coordinates of each grid cell are ( , Then the boundary of the room's outer rectangle can be obtained by statistically analyzing the range of grid coordinates: in: , These represent the minimum and maximum grid coordinates of the room in the X direction, respectively. , These represent the minimum and maximum grid coordinates of the room in the Y direction, respectively.
[0032] After obtaining the raster boundaries, the actual geometric dimensions of the room can be calculated based on the raster map resolution. in: Indicates the actual width of the room; Indicates the actual height of the room; This indicates the resolution of the raster map, which is the side length of each raster in the actual environment.
[0033] This circumscribed rectangle can not only represent the room's spatial range, but also be used for subsequent sweep direction determination and room adjacency calculation.
[0034] Step S3: Establishing Room Adjacency Relationships Let the circumscribed rectangles of the two rooms be: Room A: ( , , , ) Room B: ( , , , ) The minimum clearance between two rooms in the X direction can be calculated as follows: The minimum clearance between two rooms in the Y direction can be calculated as follows: in: This represents the minimum horizontal distance between two rooms. This represents the minimum vertical distance between two rooms.
[0035] When both conditions are met: and At that time, the two rooms can be considered to be spatially adjacent.
[0036] in: The preset adjacency threshold is used; This threshold is used to represent the maximum allowable passageway distance between two rooms.
[0037] After constructing the adjacency graph, a depth-first traversal is performed on the adjacency graph, starting from the room where the grinder begins, to obtain the room visit order.
[0038] If the room adjacency graph is not connected, then unvisited rooms are added to the end of the access order according to the minimum gap between their bounding rectangles and the bounding rectangles of the set of visited rooms, ensuring that all rooms can be visited.
[0039] Step S4: Generating the sweep path inside the room After determining the room access order, an overlay path needs to be generated within each room.
[0040] To reduce the number of turns in the path, this embodiment automatically selects the sweeping direction based on the aspect ratio of the room's circumscribed rectangle: when ≥ At that time, a horizontal sweep path is generated; when < At that time, a vertical sweep path is generated.
[0041] To ensure that no grinding area is missed, adjacent sweep paths need to maintain a certain overlap. In this embodiment, the sweep path spacing is calculated based on the working width of the grinding disc.
[0042] For example, line spacing can be determined as follows: in: Indicates the spacing between adjacent sweep paths; Indicates the effective working width of the grinding disc; This indicates the path overlap rate.
[0043] Under normal circumstances, overlap rate It can be set between 0.1 and 0.3 to ensure coverage integrity.
[0044] After generating the sweep segment, the passability of the sweep segment needs to be checked. When the line connecting the first and last points of the sweep segment is passable, only the first and last points are retained as path points. When there are impassable grids, the sweep segment is sampled grid by grid along the sweep segment direction and the center point of the passable grid is retained as a path point.
[0045] Step S5, Path Connection After generating the sweep path within the room, the sweep segments need to be connected in the order they are visited. This embodiment uses the A* algorithm for path search.
[0046] To improve search efficiency, pathfinding is prioritized within the bounded rectangle of the rooms containing the starting and ending points. If pathfinding fails, the range restriction is removed and pathfinding is performed on the entire safety grid map.
[0047] When a pathfinding for a connection fails, the corresponding target sweep segment is added back to the tail of the sweep segment connection queue, and the connection is attempted again after subsequent path planning is completed.
[0048] Step S6: Path straightening optimization The paths generated by the A* algorithm typically consist of a large number of raster nodes, and there may be redundant nodes in the paths. In order to reduce the path length and improve the running efficiency, path straightening is required.
[0049] The specific method is as follows: check if there is an unobstructed straight path between any two points in the path. If so, delete the intermediate nodes to obtain a simpler path.
[0050] Line connectivity can be detected using the Bresenham line algorithm.
[0051] Step S7: Path geometric smoothing To make the generated path more consistent with the kinematic constraints of the grinding machine, the inflection points of the path need to be smoothed.
[0052] Let the three consecutive path points be: , , Path corners can be calculated using the angle between vectors: in: Indicates the turning point of the path; Represents the path vector of the previous segment; This represents the path vector for the next segment.
[0053] Inflection point thinning is determined by calculating the angle between two adjacent path segments and the vertical distance from the path point to the line connecting the preceding and following path segments. When the angle is less than a preset angle threshold and the vertical distance is less than a preset distance threshold, the path point is deleted.
[0054] After completing the inflection point thinning and collinear point compression, an arc transition segment is inserted at the retained inflection point according to the minimum turning radius of the grinder to further optimize the path smoothness and ensure that the path meets the kinematic constraints of the grinder.
[0055] The above methods can reduce unnecessary turning points in the path, making the path smoother.
[0056] Step S8, Anti-collision verification After completing the path optimization, collision avoidance verification needs to be performed on the final path.
[0057] The specific method is as follows: convert the world coordinates of the path points into raster coordinates according to the raster map resolution and the origin coordinates, and check whether the corresponding raster is passable. If there are impassable points, delete the corresponding path segment and re-execute the path connection calculation in step S5.
[0058] The above steps can yield a complete operating path that covers the entire passable area, satisfies the motion constraints of the grinding machine, and has a safety margin.
[0059] For example, in an indoor work environment with dimensions of 40m × 30m: After generating a path using the method of this invention, automatic coverage of all passable areas can be achieved while ensuring a safe distance.
[0060] Through the above technical solution, the present invention can: Structuralized region decomposition of complex environments; Automatically plan the access order of multiple rooms; Generate high-coverage sweep paths; Optimize path smoothness; Improve path security and executability; This significantly improves the efficiency and stability of automatic operation of the grinding machine.
[0061] The above specific embodiments are merely several further embodiments of the present invention. Based on the technical solutions of the present invention and the relevant teachings of the above embodiments, those skilled in the art can make various alternative improvements and combinations to the above specific embodiments.
[0062] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, it is intended that all variations falling within the meaning and scope of equivalents of the claims be included within the present invention.
Claims
1. A method for full-area coverage path planning for a grinding machine, characterized in that... This includes the following steps: Step S1: Map acquisition and preprocessing. Acquire a grid map containing passable areas and obstacles, and inflate the obstacles in the grid map according to the equivalent radius of the grinder to obtain a safety grid map; Step S2: Room decomposition and representation. Connected components are marked for the passable areas in the safety grid map. Each connected component is treated as a room, and a bounding rectangle is determined for each room as its geometric representation. Step S3: Room access order planning. Construct a room adjacency graph based on the positional relationship between the bounding rectangles of each room, and traverse the room adjacency graph starting from the room where the grinding machine originates to determine the room access order; Step S4: Generating sweeping paths inside the room, processing each room sequentially according to the access order, determining the sweeping direction based on the aspect ratio of the room's circumscribed rectangle, and generating parallel sweeping segments covering the passable area of the room according to a preset row spacing related to the grinding width of the grinder; Step S5: Path connection, using the A* algorithm to calculate the connection paths between adjacent sweep segments and between different rooms on the security grid map; Step S6: Path straightening optimization. The straight connectivity of the connecting path is checked, and the path is straightened and simplified by finding the farthest point in the path sequence that is directly connected to the current path point. The straight connectivity is verified for drivability using the Bresenham straight line algorithm. Step S7: Path geometric smoothing, compressing collinear points in the path, thinning out small-angle inflection points, and inserting circular arc transition segments at the retained inflection points according to the minimum turning radius of the grinder to generate an executable path that satisfies kinematic constraints; Step S8: Collision avoidance verification. Perform drivability detection on the smoothed path to ensure that all points on the path are within the drivable area of the safety grid map, and output a full-area coverage path for the grinder that can cover the drivable area in the safety grid map.
2. The method for full-area coverage path planning of a grinding machine according to claim 1, characterized in that... The circumscribed rectangle is an axis-aligned rectangle. Its world coordinates are obtained by statistically analyzing the grid row and column ranges of the corresponding connected components and combining the grid map resolution with the origin coordinate transformation. The resolution of the grid map is set according to the size of the grinder and the positioning accuracy, so that the side length of a single grid is not greater than the minimum safe gap of the grinder.
3. The method for full-area coverage path planning of a grinding machine according to claim 1, characterized in that... When constructing the room adjacency graph, calculate the minimum gap between the bounding rectangles of any two rooms in the X-axis and Y-axis directions. When the minimum gap in both directions is less than or equal to the preset adjacency threshold, the two rooms are determined to be adjacent.
4. The method for full-area coverage path planning of a grinding machine according to claim 3, characterized in that... The room adjacency graph is traversed using a depth-first traversal method. When the room adjacency graph is not connected, untraversed rooms are added to the end of the access order according to the minimum gap between their bounding rectangles and the bounding rectangles of the traversed room set.
5. The method for full-area coverage path planning of a grinding machine according to claim 1, characterized in that... When the width of the circumscribed rectangle of the room is greater than or equal to its height, a sweep segment extending horizontally is generated; when the width of the circumscribed rectangle is less than its height, a sweep segment extending vertically is generated.
6. The method for full-area coverage path planning of a grinding machine according to claim 5, characterized in that... After generating the sweep segment, the passability of the sweep segment is checked. When the line connecting the first and last points of the sweep segment is passable, only the first and last points are retained as path points. When there are impassable grids, grids are sampled one by one along the sweep segment direction and the center point of the passable grid is retained as a path point.
7. The method for full-area coverage path planning of a grinding machine according to claim 1, characterized in that... When using the A* algorithm for pathfinding, priority is given to pathfinding within the combined bounding rectangle of the rooms containing the starting and ending points. If pathfinding fails, the range restriction is removed and pathfinding is performed on the entire safety grid map.
8. A method for full-area coverage path planning of a grinding machine according to claim 7, characterized in that... When a pathfinding for a connection fails, the corresponding target sweep segment is added back to the tail of the sweep segment connection queue, and the connection is attempted again after subsequent path planning is completed.
9. A method for full-area coverage path planning of a grinding machine according to claim 1, characterized in that... Inflection point thinning is determined by calculating the angle between two adjacent path segments and the vertical distance from the path point to the line connecting the preceding and following path segments. When the angle is less than a preset angle threshold and the vertical distance is less than a preset distance threshold, the path point is deleted.
10. A method for full-area coverage path planning of a grinding machine according to claim 9, characterized in that... Collision avoidance check converts the world coordinates of path points to raster coordinates based on the raster map resolution and origin coordinates, and checks whether the corresponding raster is passable. If there are impassable points, the corresponding path segment is deleted and the path connection calculation in step S5 is re-executed.