A motion control system and method for a floor-sweeping robot with multi-dimensional data fusion

Abstract

This application relates to the technical field of floor-sweeping robots, and in particular to a motion control system and method for a floor-sweeping robot with multi-dimensional data fusion. The method includes: obtaining environmental perception information, analyzing the environmental perception information, and determining a set of temporary obstacle information and a set of normal obstacle information; obtaining a set of robot parameters, and based on the set of robot parameters, according to the set of temporary obstacle information and the set of normal obstacle information, determining the cleaning dead zone of the floor-sweeping robot in the case of obstacle avoidance; based on the set of robot parameters, the environmental perception information, and the set of temporary obstacle information, analyzing the cleaning dead zone and determining a dead zone elimination strategy; according to the dead zone elimination strategy, controlling the floor-sweeping robot to perform corresponding cleaning movements, and determining and outputting the target area of the temporary obstacle. This application can ensure the cleaning effect of the floor-sweeping robot under the dual influence of the characteristics of ground obstacles and the characteristics of the space environment when the floor-sweeping robot is in a low and narrow space.

Description

Technical Field

[0001] The present application relates to the technical field of sweeping robots, and in particular to a sweeping robot motion control system and method with multi-dimensional data fusion. Background Art

[0002] In recent years, sweeping robots, as an important part of smart homes, have received widespread attention. With the development of technology, the navigation and obstacle avoidance capabilities of sweeping robots have been continuously improved. Existing high-end sweeping robots combine lidar navigation with visual navigation to improve their obstacle avoidance and scene adaptability.

[0003] However, when the sweeping robot is located in a low and narrow space, under the dual influence of the characteristics of ground obstacles and the characteristics of the spatial environment, the existing technology still cannot guarantee the cleaning effect of the sweeping robot when avoiding obstacles, resulting in the continuous retention of obstacles and dirt in the low and narrow space. Summary of the invention

[0004] The present application provides a sweeping robot motion control system and method with multi-dimensional data fusion to solve the above-mentioned technical problems.

[0005] In a first aspect, the present application provides a sweeping robot motion control method with multi-dimensional data fusion, the method comprising:

[0006] Acquire environmental perception information, analyze the environmental perception information, and determine a temporary obstacle information set and a normal obstacle information set; acquire a robot parameter set, and determine a cleaning dead zone of the sweeping robot in an obstacle avoidance situation based on the robot parameter set, the temporary obstacle information set and the normal obstacle information set; analyze the cleaning dead zone based on the robot parameter set, the environmental perception information and the temporary obstacle information set, and determine a dead zone elimination strategy; according to the dead zone elimination strategy, control the sweeping robot to perform a corresponding cleaning movement, and determine and output a temporary obstacle target area.

[0007] Through this solution, environmental perception information is analyzed to determine the temporary obstacle information set and normal obstacle information set that reflect the characteristics of temporary obstacles and normal obstacles in the area to be cleaned. On this basis, combined with the robot parameter set, the cleaning dead zone faced by the sweeping robot under the dual influence of temporary obstacles and normal obstacles is determined. For the cleaning dead zone, a dead zone elimination strategy for eliminating the cleaning dead zone is formulated. According to the dead zone elimination strategy, the sweeping robot is controlled to perform corresponding cleaning movements, and the temporary obstacle target area is determined and output. This ensures the cleaning effect of the sweeping robot under the dual influence of the characteristics of ground obstacles and the characteristics of the spatial environment when the sweeping robot is in a low and narrow space, prevents temporary obstacles and dirt from continuing to stay in the low and narrow space, and significantly improves the adaptability and cleaning coverage of the sweeping robot.

[0008] Optionally, the environmental perception information includes map data, real-time robot position data, real-time point cloud data, and real-time visual data. Analyzing the environmental perception information to determine the temporary obstacle information set and the normal obstacle information set includes:

[0009] Analyze the real-time visual data based on the map data and the real-time robot position data to determine the edge features of normal obstacles; based on the edge features of normal obstacles, analyze the real-time point cloud data to determine the center point coordinates and interference radii of normal obstacles, and determine the normal obstacle information set according to the edge features and interference radii of normal obstacles; analyze the real-time point cloud data and the real-time visual data to determine the key analysis area for temporary obstacles; adjust the lidar scanning strategy according to the key analysis area for temporary obstacles, and determine the point cloud information of temporary obstacles according to the adjusted lidar scanning strategy; determine the type of temporary obstacles, the center point coordinates of temporary obstacles, the interference radii of temporary obstacles, and the edge features of temporary obstacles according to the point cloud information of temporary obstacles and the real-time visual data, so as to determine the temporary obstacle information set.

[0010] Through this solution, based on the map data and the real-time robot position data, the real-time visual data is analyzed, and on this basis, combined with the real-time point cloud data, the center point coordinates and interference radii of normal obstacles are obtained, so as to construct the normal obstacle information set, providing a clear absolute boundary for the subsequent analysis process of the cleaning dead zone. By analyzing the real-time point cloud data and the real-time visual data, the key analysis area for temporary obstacles is determined, and through the corresponding lidar scanning strategy, the point cloud information of temporary obstacles is analyzed, so as to determine the type of temporary obstacles, the center point coordinates of temporary obstacles, the interference radii of temporary obstacles, and the edge features of temporary obstacles, so as to construct the temporary obstacle information set, realizing the accurate identification of temporary obstacles and providing a clear relative boundary for the subsequent analysis process of the cleaning dead zone.

[0011] Optionally, analyzing the real-time point cloud data and the real-time visual data to determine the key analysis area for temporary obstacles includes: analyzing the real-time visual data to determine the local color features and local edge features of temporary obstacles; analyzing the real-time point cloud data to determine the local geometric features of temporary obstacles, and performing feature fusion on the local color features, local edge features, and local geometric features of temporary obstacles to determine the comprehensive features of temporary obstacles; performing clustering processing on the comprehensive features of temporary obstacles according to a preset clustering algorithm to determine the fuzzy area of temporary obstacles; analyzing the fuzzy area of temporary obstacles to determine the center point coordinates of the fuzzy area, and determining the key analysis area for temporary obstacles according to a preset scanning radius and the center point coordinates of the fuzzy area, specifically as the following formula:

[0012] ;

[0013] Wherein, is the key analysis area of the temporary obstacle, is the coordinate of the candidate scanning point, is the abscissa of the central point coordinate of the fuzzy area, is the ordinate of the central point coordinate of the fuzzy area, is the preset scanning radius.

[0014] Through this solution, real-time visual data is analyzed to obtain the local color features and local edge features of the temporary obstacle. At the same time, by analyzing the real-time point cloud data, the local geometric features of the temporary obstacle are obtained. And through a preset clustering algorithm, clustering processing is performed on the comprehensive features of the temporary obstacle after the above features are fused to determine the fuzzy area of the temporary obstacle. On this basis, combined with the preset scanning radius and the central point coordinates of the fuzzy area, through a clear mathematical formula, the key analysis area of the temporary obstacle is scientifically quantified, improving the recognition accuracy of the temporary obstacle, and at the same time providing scientific data reference for the targeted adjustment of the lidar scanning strategy.

[0015] Optionally, the robot parameter set includes the working width and the safe collision distance. Based on the robot parameter set, the environmental perception information, and the temporary obstacle information set, analyzing the cleaning dead zone to determine the dead zone elimination strategy and the temporary obstacle target area includes:

[0016] According to the normal obstacle edge feature and the temporary obstacle edge feature, determining the area to be analyzed and the obstacle area; based on a preset analysis point division rule, dividing the area to be analyzed to determine the analysis point coordinate set; based on the working width and the safe collision distance, according to the normal obstacle edge feature, the temporary obstacle edge feature, and the analysis point coordinate set, determining the obstacle radiation area; according to the obstacle area and the obstacle radiation area, determining the cleaning dead zone.

[0017] Through this solution, according to the normal obstacle edge feature and the temporary obstacle edge feature, the area to be analyzed and the obstacle area are divided, and the area to be analyzed is divided into analysis points to determine the analysis point coordinate set. On this basis, combined with the working width and the safe collision distance corresponding to the sweeping robot, the obstacle radiation area is analyzed, and the obstacle area and the obstacle radiation area are comprehensively analyzed to determine the cleaning dead zone, improving the analysis efficiency and accuracy of the cleaning dead zone, and further improving the scientificity and comprehensiveness of the subsequent motion strategy analysis process for the cleaning dead zone.

[0018] Optionally, based on the working width and the safe collision distance, and according to the normal obstacle edge features, the temporary obstacle edge features, and the set of analysis point coordinates, determine the obstacle radiation area, specifically as the following formula:

[0019] ;

[0020] where, is the obstacle radiation area, is the current analysis point coordinate, is the th edge of the temporary obstacle, is the coordinate distance function, is the th edge of the normal obstacle, is the working width, is the safe collision distance; The determination of the cleaning dead zone according to the obstacle area and the obstacle radiation area is specifically as the following formula: where, is the cleaning dead zone, is the obstacle radiation area, is the obstacle area.

[0021] Through this solution, based on the working width and the safe collision distance, according to the normal obstacle edge features, the temporary obstacle edge features, and the set of analysis point coordinates, through explicit mathematical processing means, automatically screen out all the analysis points belonging to the obstacle radiation area, and then clarify the obstacle radiation area. On this basis, use mathematical processing means to take the union between the obstacle radiation area and the obstacle area as the cleaning dead zone, so as to accurately describe the area range of the cleaning dead zone and improve the accuracy and calculation efficiency of the cleaning dead zone determination process.

[0022] Optionally, the robot parameter set further includes the robot radius. Analyze the cleaning dead zone based on the robot parameter set and the temporary obstacle information set to determine the dead zone elimination strategy, including: analyze the temporary obstacle edge features based on the temporary obstacle type to determine the movable end point coordinates of the temporary obstacle; determine the robot travel path according to the robot real-time position data and the movable end point coordinates of the temporary obstacle; determine the moving offset and the moving direction according to the robot travel path, and thereby determine the moved active end point coordinates; determine the safe area based on the robot radius and the normal obstacle interference radius according to the temporary obstacle center point coordinates and the normal obstacle edge features; integrate the robot travel path, the moved active end point coordinates, and the safe area to determine the dead zone elimination strategy.

[0023] Through this solution, according to the type of temporary obstacle and the edge characteristics of the temporary obstacle, the coordinates of the movable end point of the temporary obstacle are determined, and combined with the real-time position data of the robot, the traveling path of the robot is planned, fully considering the different movable conditions of the temporary obstacle. On this basis, according to the traveling path of the robot, the moving offset and the moving direction are determined, and the coordinates of the movable end point after moving are obtained. At the same time, combining the radius of the robot and the interference radius of the normal obstacle, a safety area is delimited, and by integrating the traveling path of the robot, the coordinates of the movable end point after moving and the safety area, a dead zone elimination strategy is determined, so that the dead zone elimination strategy can avoid the collision between the sweeping robot and the normal obstacle while eliminating the cleaning dead zone through the targeted movement of the temporary obstacle, significantly improving the cleaning effect and adaptability of the sweeping robot.

[0024] Optionally, the integration of the traveling path of the robot, the coordinates of the movable end point after moving and the safety area to determine the dead zone elimination strategy is specifically the following formula:

[0025] ;

[0026] Wherein, is the traveling path of the robot, is the real-time position data of the robot, is the center point coordinate of the temporary obstacle of the th normal obstacle, is the radius of the robot, is the th interference radius of the normal obstacle of the th normal obstacle, is the edge feature of the th normal obstacle, is the information set of the normal obstacle, is the interference radius of the temporary obstacle, is the safety area.

[0027] Through this solution, by integrating the traveling path of the robot, the coordinates of the movable end point after moving and the safety area, the limiting conditions at each stage in the process of the sweeping robot moving the temporary obstacle are clarified, and a dead zone elimination strategy is constructed, so that the sweeping robot can avoid the collision between the sweeping robot and the normal obstacle while eliminating the current cleaning dead zone, and can effectively avoid the emergence of a new cleaning dead zone.

[0028] Optionally, the integration of the traveling path of the robot, the coordinates of the movable end point after moving and the safety area is specifically the following formula: ;

[0029] Among them, is the traveling path of the robot, is the real-time position data of the robot, is the coordinate of the movable end point of the temporary obstacle, is the coordinate of the center point of the temporary obstacle of the nth normal obstacle, is the unit travel step length, is the preset maximum step length, is the interference radius of the normal obstacle of the nth normal obstacle, is the edge feature of the normal obstacle of the nth normal obstacle, is the moved coordinate of the movable end point, is the safe area, is the moving offset, is the moving direction,

[0030] Through this solution, by using mathematical analysis methods, the planning process of the robot's traveling path, the determination process of the moved coordinate of the movable end point, and the delimitation process of the safe area are respectively described, the mathematical relationships between the robot's traveling path, the moved coordinate of the movable end point, and the safe area and their respective influencing factors are clarified, the scientificity, accuracy, and comprehensiveness of the determination processes of the robot's traveling path, the moved coordinate of the movable end point, and the safe area are improved, and further the effectiveness of the dead zone elimination strategy formulated accordingly is improved.

[0031] Optionally, according to the dead zone elimination strategy, controlling the sweeping robot to perform corresponding cleaning movements and determining and outputting the target area of the temporary obstacle includes: controlling the sweeping robot to move along the traveling path of the robot according to the dead zone elimination strategy and move the temporary obstacle; when the sweeping robot moves the temporary obstacle, controlling the cleaning structure of the sweeping robot to stop working, and determining the area outside the coverage of the normal obstacle closest to the temporary obstacle as the target area of the temporary obstacle according to the edge feature of the normal obstacle, and controlling the sweeping robot to move the temporary obstacle into the target area of the temporary obstacle; after the sweeping robot moves the temporary obstacle into the target area of the temporary obstacle, controlling the sweeping robot to clean the cleaning dead zone, and recording and outputting the target area of the temporary obstacle.

[0032] Through this solution, according to the dead zone elimination strategy, the sweeping robot is controlled to move the temporary obstacle. During this process, the cleaning structure of the sweeping robot is controlled to stop working, so as to avoid the sweeping robot from malfunctioning due to the influence of the temporary obstacle during the movement. At the same time, the sweeping robot is controlled to move the temporary obstacle into the temporary obstacle target area. After the sweeping robot completes the cleaning work on the original cleaning dead zone, the temporary obstacle target area is provided to the user, so that the user can timely know the position of the temporary obstacle, facilitating the user to quickly handle the temporary obstacle and effectively avoiding the continuous retention of the temporary obstacle in the current area.

[0033] In a second aspect, the present application provides a motion control system for a sweeping robot with multi-dimensional data fusion. The system includes:

[0034] An obstacle analysis module for obtaining environmental perception information, analyzing the environmental perception information, and determining a temporary obstacle information set and a normal obstacle information set; a region analysis module for obtaining a robot parameter set, and based on the robot parameter set, according to the temporary obstacle information set and the normal obstacle information set, determining a cleaning dead zone of the sweeping robot under obstacle avoidance conditions; a motion analysis module for analyzing the cleaning dead zone based on the robot parameter set, the environmental perception information, and the temporary obstacle information set, and determining a dead zone elimination strategy; a motion control module for controlling the sweeping robot to perform corresponding cleaning motions according to the dead zone elimination strategy, determining and outputting a temporary obstacle target area. Description of the Drawings

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0036] Figure 1 It is a schematic diagram of an application scenario provided by an embodiment of the present application;

[0037] Figure 2 It is a flowchart of a motion control method for a sweeping robot with multi-dimensional data fusion provided by an embodiment of the present application;

[0038] Figure 3 It is a schematic structural diagram of a motion control system for a sweeping robot with multi-dimensional data fusion provided by an embodiment of the present application. Detailed Embodiments

[0039] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the following will clearly and completely describe the technical solutions in the embodiments of this application with reference to the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are some but not all of the embodiments of this application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of this application without creative efforts shall fall within the scope of protection of this application.

[0040] In addition, the term "and / or" in this document is merely a description of the association relationship between associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this document generally represents an "or" relationship between the associated objects before and after, unless otherwise specified.

[0041] The following further describes the embodiments of this application in detail with reference to the accompanying drawings of the specification.

[0042] However, when the sweeping robot is in a low and narrow space, under the dual influence of the characteristics of ground obstacle items and the characteristics of the space environment, the prior art still has difficulty ensuring the cleaning effect of the sweeping robot during obstacle avoidance, resulting in the continuous retention of obstacles and dirt in the low and narrow space.

[0043] Based on this, this application provides a motion control system and method for a sweeping robot with multi-dimensional data fusion. Analyze the environmental perception information to determine a set of temporary obstacle information and a set of normal obstacle information that reflect the characteristics of temporary obstacles and normal obstacles in the area to be cleaned. On this basis, combined with the set of robot parameters, determine the cleaning dead zone faced by the sweeping robot under the dual influence of temporary obstacles and normal obstacles. For the cleaning dead zone, formulate a dead zone elimination strategy for eliminating the cleaning dead zone, and according to the dead zone elimination strategy, control the sweeping robot to perform corresponding cleaning movements, determine and output the target area of the temporary obstacle, ensure the cleaning effect of the sweeping robot when it is in a low and narrow space under the dual influence of the characteristics of ground obstacle items and the characteristics of the space environment, prevent the continuous retention of temporary obstacles and dirt in the low and narrow space, and significantly improve the adaptability and cleaning coverage rate of the sweeping robot.

[0044] Figure 1 This is a schematic diagram of an application scenario provided by this application. During the operation of the sweeping robot, applying the method provided by this application can ensure the cleaning effect of the sweeping robot when it is in a low and narrow space under the dual influence of the characteristics of ground obstacle items and the characteristics of the space environment.

[0045] Specifically, the method of the present application is applied to any server, which can be installed inside a floor cleaning robot. The server obtains and analyzes the environmental perception information provided by the floor cleaning robot, determines a set of temporary obstacle information and a set of normal obstacle information reflecting the characteristics of temporary obstacles and normal obstacles in the area to be cleaned. On this basis, combined with the robot parameter set, the cleaning dead zones faced by the floor cleaning robot under the dual influence of temporary obstacles and normal obstacles are determined. For the cleaning dead zones, a dead zone elimination strategy for eliminating the cleaning dead zones is formulated, and according to the dead zone elimination strategy, the floor cleaning robot is controlled to perform corresponding cleaning movements, determine and output the target area of the temporary obstacles, ensuring that when the floor cleaning robot is in a low and narrow space, the cleaning effect under the dual influence of the characteristics of ground obstacle items and the characteristics of the space environment is ensured, preventing temporary obstacles and dirt from continuously staying in the low and narrow space, and significantly improving the adaptability and cleaning coverage rate of the floor cleaning robot. The specific implementation manner can refer to the following embodiments.

[0046] Figure 2 The flowchart of a method for controlling the movement of a multi-dimensional data fusion floor cleaning robot provided by an embodiment of the present application. The method of this embodiment can be applied to the server in the above scenario. As Figure 2 shown, the method includes:

[0047] S201. Obtain environmental perception information, analyze the environmental perception information, and determine a set of temporary obstacle information and a set of normal obstacle information.

[0048] The environmental perception information can be the information of the home environment where the floor cleaning robot is located. The environmental perception information can be obtained by the cooperation of a lidar sensor and a vision sensor built in the floor cleaning robot. The set of temporary obstacle information can be a set of information corresponding to the characteristics of temporary obstacles existing in the current area to be cleaned faced by the floor cleaning robot. The temporary obstacles can be obstacles that temporarily appear in the area and can be moved, such as small toys, data cables, etc. The set of normal obstacle information can be a set of information corresponding to the characteristics of normal obstacles existing in the current area to be cleaned faced by the floor cleaning robot. The set of normal obstacle information can be obstacles that normally exist in the area and are usually immovable, such as furniture, walls, etc.

[0049] Specifically, when a sweeping robot is cleaning a low and narrow area, such as the bottom of a sofa, if there are temporary obstacles on the ground, the accuracy of the sweeping robot's obstacle recognition is easily affected by both the surface characteristics of the object and environmental factors. The surface characteristics of the object mainly affect the lidar's recognition of obstacles. Obstacles with weak surface reflection characteristics, such as data cables with rubber insulating materials, have weak echo signals on their surfaces for laser pulses, making it difficult for the lidar to fully and accurately recognize them. Environmental factors mainly affect the visual sensor's recognition of obstacles. The light conditions in low and narrow areas are poor, and the sweeping robot's own light source range is limited and easily affected by other objects with high light reflectivity in the surrounding area. Therefore, the recognition of obstacles in low and narrow areas requires complementary scheduling of the lidar sensor and the visual sensor to effectively analyze environmental perception information and correctly identify temporary obstacles and normal obstacles.

[0050] After identifying temporary obstacles and normal obstacles, existing sweeping robots will avoid them through their built-in obstacle avoidance algorithms. However, this method is likely to cause inadequate cleaning in low and narrow areas, because in low and narrow areas, temporary obstacles are likely to form a cleaning dead zone together with normal obstacles. The cleaning dead zone does not require the temporary obstacles and normal obstacles to form a closed area, but only requires the interval between the temporary obstacles and the normal obstacles to be smaller than the width through which the sweeping robot can pass. Especially in low and narrow areas, since the space available for the sweeping robot to adjust its movement is relatively small, Small, temporary obstacles of small size can form a large cleaning dead zone with the surrounding normal obstacles, and low and narrow areas are usually easily ignored by users, which causes temporary obstacles to remain in the area for a long time, resulting in the dirt in the corresponding cleaning dead zone cannot be cleaned for a long time, which in turn has a negative impact on the user's home environment. Therefore, after identifying temporary obstacles and normal obstacles in low and narrow areas, the characteristics of temporary obstacles and normal obstacles reflected in the environmental perception information are extracted to form temporary obstacle information sets and normal obstacle information sets, which provide a data basis for the subsequent analysis of cleaning dead zones and the formulation of corresponding cleaning strategies.

[0051] S202: Obtain a robot parameter set, and determine a cleaning dead zone of the sweeping robot in an obstacle avoidance situation based on the robot parameter set, a temporary obstacle information set, and a normal obstacle information set.

[0052] The robot parameter set may be a set of device parameters corresponding to the sweeping robot, such as volume, radius, etc., and the robot parameter set may be provided by the sweeping robot. The cleaning dead zone may be an area formed by temporary obstacles and normal obstacles in a low and narrow area that is difficult to be directly cleaned by the sweeping robot.

[0053] Specifically, the concept of the cleaning dead zone is related not only to the temporary obstacles and normal obstacles that directly constitute the cleaning dead zone, but also to the specific parameters of the floor cleaning robot. For the same area composed of temporary obstacles and normal obstacles, the size and shape of the cleaning dead zone faced by floor cleaning robots with different volumes and passabilities will vary. It is necessary to comprehensively analyze the temporary obstacle information set and the normal obstacle information set based on the robot parameter set through mathematical analysis means to quantitatively obtain the cleaning dead zone formed by the floor cleaning robot taking obstacle avoidance measures under the dual influence of the current temporary obstacles and normal obstacles, providing scientific data support for the subsequent analysis process of eliminating the cleaning dead zone.

[0054] S203. Analyze the cleaning dead zone based on the robot parameter set, environmental perception information, and temporary obstacle information set, and determine the dead zone elimination strategy.

[0055] The dead zone elimination strategy can be a control strategy aimed at eliminating the cleaning dead zone and used to control the floor cleaning robot to perform corresponding movements.

[0056] Specifically, after determining the cleaning dead zone faced by the current floor cleaning robot, it is necessary to plan the subsequent movement trajectory of the floor cleaning robot to achieve the cleaning of the cleaning dead zone. Since the position of the normal obstacle cannot change, in order to let the floor cleaning robot enter the cleaning area for cleaning work, it is necessary to change the position of the temporary obstacle. Through mathematical analysis means, based on the robot parameter set, environmental perception information, and temporary obstacle information set, analyze the cleaning dead zone, and plan the movement strategy of the floor cleaning robot to move the temporary obstacle so that the changed temporary obstacle can no longer form an inaccessible cleaning dead zone with the corresponding normal obstacle, and at the same time avoid the moved temporary obstacle forming a new cleaning dead zone with other normal obstacles, realizing the elimination of the cleaning dead zone to improve the cleaning effect of the floor cleaning robot.

[0057] S204. According to the dead zone elimination strategy, control the floor cleaning robot to perform corresponding cleaning movements, and determine and output the target area of the temporary obstacle.

[0058] The target area of the temporary obstacle can be the final area where the corresponding temporary obstacle needs to be moved after the floor cleaning robot completes the corresponding cleaning work.

[0059] Specifically, after determining the dead zone elimination strategy, according to the corresponding travel route in the dead zone elimination strategy, control the sweeping robot to move the temporary obstacle, and after eliminating the cleaning dead zone, clean the cleaning area that could not be entered originally. After the cleaning is completed, further move the temporary obstacle outside the current low and narrow area, mark the area where the temporary obstacle finally locates as the temporary obstacle target area, and provide this area to the user, so that the user can clean the temporary obstacle in time, avoid the temporary obstacle from affecting the subsequent cleaning work of the sweeping robot, and at the same time help the user find small items lost in the corner.

[0060] Through this solution, analyze the environmental perception information, determine the temporary obstacle information set and the normal obstacle information set reflecting the characteristics of temporary obstacles and normal obstacles in the area to be cleaned, and on this basis, combine with the robot parameter set to determine the cleaning dead zone faced by the sweeping robot under the dual influence of temporary obstacles and normal obstacles. For the cleaning dead zone, formulate a dead zone elimination strategy for eliminating the cleaning dead zone, and according to the dead zone elimination strategy, control the sweeping robot to perform corresponding cleaning movements, determine and output the temporary obstacle target area, ensure the cleaning effect of the sweeping robot under the dual influence of the characteristics of ground obstacle items and the spatial environment characteristics when in a low and narrow space, prevent temporary obstacles and dirt from staying continuously in the low and narrow space, and significantly improve the adaptability and cleaning coverage rate of the sweeping robot.

[0061] In some embodiments, based on the map data and the real-time position data of the robot, analyze the real-time visual data to determine the edge characteristics of normal obstacles; based on the edge characteristics of normal obstacles, analyze the real-time point cloud data to determine the center point coordinates and the interference radius of normal obstacles, and according to the edge characteristics of normal obstacles and the interference radius of normal obstacles, determine the normal obstacle information set; analyze the real-time point cloud data and the real-time visual data to determine the key analysis area of temporary obstacles; according to the key analysis area of temporary obstacles, adjust the lidar scanning strategy, and according to the adjusted lidar scanning strategy, determine the point cloud information of temporary obstacles; according to the point cloud information of temporary obstacles and the real-time visual data, determine the type of temporary obstacles, the center point coordinates of temporary obstacles, the interference radius of temporary obstacles and the edge characteristics of temporary obstacles, so as to determine the temporary obstacle information set.

[0062] The environmental perception information includes map data, real-time robot position data, real-time point cloud data, and real-time visual data. The map data can be the map information of the user's home interior constructed by the lidar sensor built into the floor cleaning robot. The real-time robot position data can be the position coordinates of the floor cleaning robot within the user's home during operation. The real-time point cloud data can be the local three-dimensional space data of the environment where the floor cleaning robot is located, collected by the lidar built into the floor cleaning robot. The real-time visual data can be the local visual data of the environment where the floor cleaning robot is located, collected by the visual sensor built into the floor cleaning robot. The normal obstacle edge feature can be the information representing the local edge feature corresponding to the normal obstacle affecting the current floor cleaning robot. The normal obstacle center point coordinates can be the local center point coordinates of the normal obstacle affecting the current floor cleaning robot. The normal obstacle interference radius can be the local influence radius of the normal obstacle affecting the current floor cleaning robot. The temporary obstacle key analysis area can be the area where the temporary obstacle is located. The lidar scanning strategy can be the operating strategy of the lidar built into the floor cleaning robot, which can include scanning angle, scanning frequency, and scanning area, etc. The temporary obstacle point cloud information can be the point cloud information used to represent the shape characteristics of the temporary obstacle. The temporary obstacle type can be the item type to which the temporary obstacle belongs. The temporary obstacle center point coordinates can be the center point coordinates of the temporary obstacle in the current state. The temporary obstacle interference radius can be the influence radius of the temporary obstacle on the floor cleaning robot. The temporary obstacle edge feature can be the information used to represent the edge feature of the current temporary obstacle.

[0063] Specifically, during the process of analyzing normal obstacles, since normal obstacles are usually relatively large in volume compared to the floor cleaning robot, it is usually not the entire normal obstacle but a local part of the normal obstacle that constitutes the cleaning dead zone with the temporary obstacle. The distribution position of the normal obstacles in the home can be reflected from the map data pre-collected by the floor cleaning robot. Combining with the real-time robot position data, it can be obtained which relative position of which normal obstacles the current floor cleaning robot is located at. Furthermore, through the edge recognition technology in image analysis technology, the real-time visual data is analyzed to extract the normal obstacle edge features within the cleaning area to be cleaned faced by the current floor cleaning robot. The normal obstacle edge features can provide a clear absolute boundary for the subsequent analysis process of the cleaning dead zone. Based on the normal obstacle edge features, the point cloud data located within the normal obstacle edge is screened from the real-time point cloud data. These point cloud data can jointly reflect the geometric features of the normal obstacle. Based on this, the normal obstacle center point coordinates and the normal obstacle interference radius are obtained to reflect the maximum influence range of the normal obstacle. The normal obstacle information set thus constituted can provide scientific guiding data for the subsequent analysis of the floor cleaning robot's motion strategy.

[0064] After obtaining the normal obstacle information set through analysis, it is necessary to analyze the temporary obstacles. From the description of the foregoing embodiments, it can be seen that the accurate analysis of temporary obstacles requires the good cooperation of lidar and visual sensors. Through mathematical analysis means, based on the real-time point cloud data and real-time visual data, the key analysis area of the temporary obstacle is delimited. This area provides guidance for the further data collection of the lidar. According to the relative angle, relative distance between the key analysis area of the temporary obstacle and the current position of the sweeping robot, and the size of the key analysis area of the temporary obstacle, the scanning angle, scanning frequency and scanning range of the lidar are adjusted respectively, so that the lidar changes from the originally default ring scanning strategy with wide adaptability to a special scanning strategy for temporary obstacles, improving the recognition accuracy of temporary obstacles. Through the method of using the visual sensor to assist in determining the key scanning area and using the lidar for targeted scanning, the good cooperation between the lidar and the visual sensor is achieved, and the temporary obstacle point cloud information that can accurately reflect the characteristics of the temporary obstacle is obtained. On this basis, through the entity recognition algorithm, the type of the temporary obstacle is obtained, and the obstacle center point coordinates, the interference radius of the temporary obstacle and the edge characteristics of the temporary obstacle are screened from the obstacle point cloud information, so as to construct the temporary obstacle information set, providing a clear relative boundary for the subsequent analysis process of the cleaning dead zone.

[0065] Through this solution, based on the map data and the real-time position data of the robot, the real-time visual data is analyzed. On this basis, combined with the real-time point cloud data, the center point coordinates of the normal obstacle and the interference radius of the normal obstacle are obtained, so as to construct the normal obstacle information set, providing a clear absolute boundary for the subsequent analysis process of the cleaning dead zone. By analyzing the real-time point cloud data and the real-time visual data, the key analysis area of the temporary obstacle is determined, and through the corresponding lidar scanning strategy, the temporary obstacle point cloud information is analyzed, so as to determine the type of the temporary obstacle, the center point coordinates of the temporary obstacle, the interference radius of the temporary obstacle and the edge characteristics of the temporary obstacle, so as to construct the temporary obstacle information set, realizing the accurate recognition of the temporary obstacle and providing a clear relative boundary for the subsequent analysis process of the cleaning dead zone.

[0066] In some embodiments, the real-time visual data is analyzed to determine the local color characteristics and local edge characteristics of the temporary obstacle; the real-time point cloud data is analyzed to determine the local geometric characteristics of the temporary obstacle, and the local color characteristics, local edge characteristics and local geometric characteristics of the temporary obstacle are feature-fused to determine the comprehensive characteristics of the temporary obstacle; according to the preset clustering algorithm, the comprehensive characteristics of the temporary obstacle are clustered to determine the fuzzy area of the temporary obstacle; the fuzzy area of the temporary obstacle is analyzed to determine the center point coordinates of the fuzzy area, and according to the preset scanning radius and the center point coordinates of the fuzzy area, the key analysis area of the temporary obstacle is determined, specifically as the following formula (1):

[0067] (1)

[0068] Wherein, is the key analysis area of the temporary obstacle, is the coordinate of the candidate scanning point, is the abscissa of the center point coordinate of the fuzzy area, is the ordinate of the center point coordinate of the fuzzy area, is the preset scanning radius.

[0069] The local color feature of the temporary obstacle can be the information used to characterize the local color feature of the temporary obstacle. The local edge feature of the temporary obstacle can be the information used to characterize the local edge feature of the temporary obstacle. The local geometric feature of the temporary obstacle can be the information used to characterize the local geometric feature of the temporary obstacle. The comprehensive feature of the temporary obstacle can be the data obtained by splicing the local color feature, local edge feature and local geometric feature corresponding to the temporary obstacle through the feature splicing algorithm, and is used to characterize the comprehensive feature of the temporary obstacle. The preset clustering algorithm can be a preset algorithm for classifying and distinguishing the comprehensive feature of the temporary obstacle from other features in the area. The preset clustering algorithm can adopt existing mature clustering algorithms such as the K-means algorithm and the DBSCAN algorithm. The fuzzy area of the temporary obstacle can be the area where the temporary obstacle is initially determined. The center point coordinate of the fuzzy area can be the center point coordinate corresponding to the fuzzy area of the temporary obstacle. The preset scanning radius can be the maximum scanning radius supported by the lidar built in the sweeping robot.

[0070] Specifically, due to the influence of problems such as perspective limitations and insufficient light on the real-time visual data collected by the sweeping robot through the visual sensor, it is difficult to reflect the complete characteristics of the temporary obstacle. Through the color histogram analysis algorithm and edge detection algorithm in the image analysis technology, the local color characteristics and local edge characteristics of the temporary obstacle reflected in the real-time visual data are extracted respectively. The local color characteristics and local edge characteristics of the temporary obstacle can jointly reflect the boundary between the temporary obstacle and the surrounding objects and regions. Similarly, due to the influence of problems such as the default scanning strategy and the surface characteristics of objects on the real-time point cloud data collected by the sweeping robot through the lidar, it is also difficult to reflect the complete characteristics of the temporary obstacle. The real-time point cloud data can mainly reflect the local geometric structure characteristics of the temporary obstacle, and the corresponding local geometric structure characteristics of the temporary obstacle can provide guidance for the judgment of the temporary obstacle type, and then deduce the influence area of the current temporary obstacle. Through the feature vector splicing algorithm, the local color characteristics, local edge characteristics and local geometric characteristics of the temporary obstacle are fused to determine the comprehensive characteristics of the temporary obstacle, and through the preset clustering algorithm, the comprehensive characteristics of the temporary obstacle are clustered to initially distinguish the boundary of the area where the temporary obstacle is located to determine the fuzzy area of the temporary obstacle, and then the key analysis area of the temporary obstacle is scientifically quantified through formula (1) to improve the recognition accuracy of the temporary obstacle and at the same time provide scientific data reference for the targeted adjustment of the lidar scanning strategy.

[0071] Through this solution, the real-time visual data is analyzed to obtain the local color characteristics and local edge characteristics of the temporary obstacle. At the same time, by analyzing the real-time point cloud data, the local geometric characteristics of the temporary obstacle are obtained, and through the preset clustering algorithm, the comprehensive characteristics of the temporary obstacle after the above feature fusion are clustered to determine the fuzzy area of the temporary obstacle. On this basis, combined with the preset scanning radius and the coordinates of the center point of the fuzzy area, through a clear mathematical formula, the key analysis area of the temporary obstacle is scientifically quantified to improve the recognition accuracy of the temporary obstacle and at the same time provide scientific data reference for the targeted adjustment of the lidar scanning strategy.

[0072] In some embodiments, according to the edge characteristics of the normal obstacle and the edge characteristics of the temporary obstacle, the area to be analyzed and the obstacle area are determined; based on the preset analysis point division rule, the area to be analyzed is divided into analysis points to determine the analysis point coordinate set; based on the working width and the safe collision distance, according to the edge characteristics of the normal obstacle, the edge characteristics of the temporary obstacle and the analysis point coordinate set, the obstacle radiation area is determined; according to the obstacle area and the obstacle radiation area, the cleaning dead zone is determined.

[0073] The robot parameter set includes the working width and the safe collision distance. The working width can be the maximum width of the floor cleaning robot in the working state. The safe collision distance can be the preset collision warning distance inside the floor cleaning robot. The area to be analyzed can be the corresponding area where there may be cleaning dead zones under the influence of current temporary obstacles and normal obstacles. The obstacle area can be the area occupied by the identified temporary obstacles and normal obstacles. The preset analysis point division rule can be a preset rule for dividing the area to be analyzed, which can include parameters such as the density of analysis points, and the preset analysis point division rule can be set according to specific requirements. The analysis point coordinate set can be a set of the coordinate positions of all divided analysis points. The obstacle radiation area can be the area that the floor cleaning robot cannot pass through under the combined influence of temporary obstacles and normal obstacles.

[0074] Specifically, through the edge features of normal obstacles and the edge features of temporary obstacles, the area jointly formed between the edges of normal obstacles and the edges of temporary obstacles is extracted. Among them, the areas occupied by the edges of normal obstacles and the edges of temporary obstacles are the obstacle areas, and the floor cleaning robot cannot touch this area, otherwise it will have a direct collision with normal obstacles or temporary obstacles. And the area outside the areas occupied by the edges of normal obstacles and the edges of temporary obstacles is the area to be analyzed. There may be cleaning dead zones in this area and further analysis is required. Since the area to be analyzed consists of countless points, in order to balance the analysis accuracy and analysis efficiency, according to the preset analysis point division rule, the area to be analyzed is divided into analysis points, and the analysis point coordinate set is extracted to control the calculation scale. On this basis, since the judgment of cleaning dead zones mainly depends on whether the floor cleaning robot can enter this area without colliding with temporary obstacles or normal obstacles. If it can enter, then this area is not a cleaning dead zone. If it cannot enter, then this area is a cleaning dead zone. The passability of the floor cleaning robot mainly depends on its corresponding working width and safe collision distance. The working width determines whether the floor cleaning robot can smoothly pass through the designated area, and the safe collision distance determines whether the floor cleaning robot has room for movement adjustment. Through mathematical analysis means, based on the working width and safe collision distance, according to the edge features of normal obstacles, the edge features of temporary obstacles and the analysis point coordinate set, the obstacle radiation area is calculated, and by synthesizing the obstacle area and the obstacle radiation area, the cleaning dead zone is determined.

[0075] Through this solution, according to the edge features of normal obstacles and the edge features of temporary obstacles, the area to be analyzed and the obstacle area are divided, and the area to be analyzed is divided into analysis points to determine the analysis point coordinate set. On this basis, combined with the corresponding working width and safe collision distance of the floor cleaning robot, the obstacle radiation area is analyzed, and by comprehensively analyzing the obstacle area and the obstacle radiation area, the cleaning dead zone is determined, improving the analysis efficiency and accuracy of the cleaning dead zone, and further improving the scientificity and comprehensiveness of the subsequent motion strategy analysis process for the cleaning dead zone.

[0076] In some embodiments, based on the working width and the safe collision distance, according to the edge features of normal obstacles, the edge features of temporary obstacles, and the set of analysis point coordinates, the obstacle radiation area is determined, specifically by the following formula (2):

[0077] (2)

[0078] Wherein, is the obstacle radiation area, is the current analysis point coordinate, is the th edge of the temporary obstacle, is the coordinate distance function, is the th edge of the normal obstacle, is the working width, is the safe collision distance; according to the obstacle area and the obstacle radiation area, the cleaning dead zone is determined, specifically by the following formula (3): (3) Wherein, is the cleaning dead zone, is the obstacle radiation area, is the obstacle area. The coordinate distance function can be a mathematical function used to calculate the relative distance between different coordinate points.

[0079] Specifically, the obstacle radiation area is defined by formula (2), where the condition means that if there is an analysis point whose relative distance from the temporary obstacle or from the normal obstacle is less than the working width of the sweeping robot, then this analysis point belongs to the obstacle radiation area, and the condition means that if there is an analysis point whose relative distance from the temporary obstacle or from the normal obstacle is less than the safe collision distance of the sweeping robot, then this analysis point belongs to the obstacle radiation area. By automatically screening all the analysis points belonging to the obstacle radiation area through formula (2), the obstacle radiation area is further clarified. On this basis, the union between the obstacle radiation area and the obstacle area is used as the cleaning dead zone through formula (3) to accurately describe the area range of the cleaning dead zone.

[0080] Through this solution, based on the working width and the safe collision distance, according to the edge features of normal obstacles, the edge features of temporary obstacles, and the set of analysis point coordinates, through clear mathematical processing means, all the analysis points belonging to the obstacle radiation area are automatically screened, and then the obstacle radiation area is clarified. On this basis, through mathematical processing means, the union between the obstacle radiation area and the obstacle area is used as the cleaning dead zone to accurately describe the area range of the cleaning dead zone, improving the accuracy and calculation efficiency of the cleaning dead zone determination process.

[0081] In some embodiments, based on the type of temporary obstacle, analyze the edge features of the temporary obstacle to determine the coordinates of the movable endpoints of the temporary obstacle; according to the real-time position data of the robot and the coordinates of the movable endpoints of the temporary obstacle, determine the travel path of the robot; according to the travel path of the robot, determine the movement offset and movement direction, and thereby determine the coordinates of the movable endpoints after movement; based on the radius of the robot and the interference radius of the normal obstacle, according to the coordinates of the center point of the temporary obstacle and the edge features of the normal obstacle, determine the safe area; integrate the travel path of the robot, the coordinates of the movable endpoints after movement and the safe area, and thereby determine the dead zone elimination strategy.

[0082] The robot parameter set further includes the radius of the robot, and the radius of the robot can be the radius dimension of the robot body. The coordinates of the movable endpoints of the temporary obstacle can be the coordinates of the endpoints where the current temporary obstacle can be moved. The travel path of the robot can be the route of movement required for the floor cleaning robot to move the temporary obstacle. The movement offset can be the moving distance of the floor cleaning robot for the temporary obstacle. The movement direction can be the moving direction of the floor cleaning robot for the temporary obstacle. The coordinates of the movable endpoints after movement can be the position coordinates where the movable endpoints of the temporary obstacle are located after being moved by the floor cleaning robot according to the current corresponding travel path of the robot. The safe area can be a safe area where the floor cleaning robot can perform dynamic path planning without being affected by other factors.

[0083] Specifically, after determining the cleaning dead zone, it is necessary to control the sweeping robot to move the temporary obstacle to break the establishment condition of the cleaning dead zone in the foregoing embodiment, thereby eliminating the cleaning dead zone. In this process, it is first necessary to clarify how to move the temporary obstacle. Here, considering that there are two situations for the temporary obstacle: freely movable and not freely movable. Freely movable means that the temporary obstacle is not connected to other items and can be independently and freely moved, while not freely movable means that the temporary obstacle is connected to other items, such as a charging cable or a power strip wire hanging on the ground, one end of which is connected to other items, with limited range of movement and only the unconnected end can be moved. Therefore, the concept of the movable end point coordinates of the temporary obstacle is introduced to provide a position reference for the movable point during the process of the sweeping robot moving the temporary obstacle. The movable end point coordinates of the temporary obstacle are directly related to the type of the temporary obstacle and the edge characteristics of the temporary obstacle. The type of the temporary obstacle determines whether the temporary obstacle is an item that is not freely movable, and the edge characteristics of the temporary obstacle determine the position of the corresponding movable end point of the temporary obstacle. By comprehensively analyzing the type of the temporary obstacle and the edge characteristics of the temporary obstacle, the movable end point coordinates of the temporary obstacle are extracted, and through mathematical analysis means, based on the real-time position data of the robot and the movable end point coordinates of the temporary obstacle, the robot travel path required for the sweeping robot to move the temporary obstacle is planned, and the moving offset and moving direction are determined accordingly, and then the movable end point coordinates after moving are extracted. If the movable end point coordinates after moving still form a cleaning dead zone with other obstacles, it is necessary to further plan the moving path. At this time, the sweeping robot needs to be in a safe area to ensure the accuracy of the new route. Through mathematical analysis means, based on the radius of the robot and the interference radius of the normal obstacle, the corresponding range of the safe area is quantified according to the center point coordinates of the temporary obstacle and the edge characteristics of the normal obstacle. By integrating the above robot travel path, the movable end point coordinates after moving and the safe area, a dead zone elimination strategy is obtained.

[0084] Through this solution, according to the type of the temporary obstacle and the edge characteristics of the temporary obstacle, the movable end point coordinates of the temporary obstacle are determined, and combined with the real-time position data of the robot, the robot travel path is planned, fully considering the different movable conditions of the temporary obstacle. On this basis, according to the robot travel path, the moving offset and moving direction are determined, and the movable end point coordinates after moving are obtained. At the same time, combined with the radius of the robot and the interference radius of the normal obstacle, a safe area is delimited, and by integrating the robot travel path, the movable end point coordinates after moving and the safe area, a dead zone elimination strategy is determined, enabling the dead zone elimination strategy to avoid collisions between the sweeping robot and normal obstacles while eliminating the cleaning dead zone through targeted movement of the temporary obstacle, significantly improving the cleaning effect and adaptability of the sweeping robot.

[0085] In some embodiments, the robot travel path, the coordinates of the moving end point after movement, and the safety area are integrated to determine the dead zone elimination strategy, specifically the following formula (4):

[0086] (4)

[0087] Wherein, is the robot travel path, is the real-time position data of the robot, is the coordinate of the center point of the temporary obstacle of the th normal obstacle, is the radius of the robot, is the interference radius of the th normal obstacle, is the edge feature of the th normal obstacle, is the information set of normal obstacles, is the coordinate of the moving end point after movement, is the interference radius of the temporary obstacle, is the safety area.

[0088] Specifically, when the sweeping robot moves according to the robot travel route to move the temporary obstacle, it is necessary to ensure that the sweeping robot does not collide with other normal obstacles. By in formula (4), the relative distance between the sweeping robot and the normal obstacle is restricted to be greater than the sum of the corresponding interference radii of the two, that is, the sweeping robot always maintains a safe distance from each normal obstacle during the movement. After moving the temporary obstacle, it is also necessary to prevent new interference between the temporary obstacle and the normal obstacle. By the relationship between the temporary obstacle and the normal obstacle after movement is defined. If the above conditions cannot be met, it means that the travel path needs to be updated. By restricting the sweeping robot to re-plan the path within the safety area, the interference of the sweeping robot during the path planning process can be prevented. Combining the above three steps, the dead zone elimination strategy is constructed.

[0089] Through this solution, the robot travel path, the coordinates of the moving end point after movement, and the safety area are integrated, the limiting conditions at each stage during the process of the sweeping robot moving the temporary obstacle are clarified, and the dead zone elimination strategy is constructed, enabling the sweeping robot to eliminate the current cleaning dead zone while avoiding collisions between the sweeping robot and normal obstacles and effectively avoiding the emergence of new cleaning dead zones.

[0090] In some embodiments, the robot travel path, the coordinates of the moving end point after movement, and the safety area are integrated, specifically the following formula (5):

[0091] (5)

[0092] Among them, is the robot's traveling path, is the robot's real-time position data, is the coordinate of the movable end point of the temporary obstacle, is the coordinate of the center point of the temporary obstacle of the nth normal obstacle, is the unit traveling step length, is the preset maximum step length, is the radius of the robot, is the interference radius of the nth normal obstacle of the normal obstacle, is the edge feature of the nth normal obstacle of the normal obstacle, is the coordinate of the movable end point after moving, is the safety area, is the moving offset, is the moving direction, is the coordinate of the current analysis point.

[0093] The unit traveling step length can be the step length of the floor cleaning robot each time it travels under the current robot traveling path.

[0094] The preset maximum step length can be the maximum step length that the floor cleaning robot can directly travel in the current environment. The preset maximum step length is directly related to the number of obstacles around the floor cleaning robot. The preset maximum step length can be derived through the linear relationship between the traveling step length and the number of obstacles.

[0095] Specifically, the direction vector pointing from the robot's real-time position to the coordinate of the movable end point of the temporary obstacle is described by in formula (5), and the unit direction vector is described by, combined with the unit traveling step length, based on the current robot's real-time position data, plan the robot's traveling route, and through in formula (5), calculate the position of the coordinate of the movable end point of the temporary obstacle after being applied with the corresponding moving offset in the moving direction, and obtain the coordinate of the movable end point after moving. The safety area is defined by in formula (5) as consisting of all those that can maintain a safe distance from the floor cleaning robot to prevent the floor cleaning robot from colliding with normal obstacles.

[0096] ​​​​Through this scheme, mathematical analysis methods are used to describe the planning process of the robot's travel path, the determination process of the active endpoint coordinates after movement, and the demarcation process of the safe area, respectively, to clarify the mathematical relationship between the robot's travel path, the coordinates of the active endpoint after movement, and the safe area and their respective influencing factors, to improve the scientificity, accuracy and comprehensiveness of the robot's travel path, the coordinates of the active endpoint after movement, and the safe area determination process, thereby improving the effectiveness of the dead zone elimination strategy formulated based on this.

[0097] In some embodiments, according to the dead zone elimination strategy, the sweeping robot is controlled to move along the robot's travel path to move the temporary obstacle; when the sweeping robot moves the temporary obstacle, the cleaning structure of the sweeping robot is controlled to stop working, and according to the edge characteristics of the normal obstacle, the area outside the coverage range of the normal obstacle closest to the temporary obstacle is determined as the temporary obstacle target area, and the sweeping robot is controlled to move the temporary obstacle into the temporary obstacle target area; after the sweeping robot moves the temporary obstacle into the temporary obstacle target area, the sweeping robot is controlled to clean the cleaning dead zone, and the temporary obstacle target area is recorded and output.

[0098] Specifically, according to the dead zone elimination strategy, the built-in microcontroller of the sweeping robot is used to control the sweeping robot to move along the robot's travel path to achieve the removal of temporary obstacles. During the movement process, the cleaning structure on the sweeping robot, such as the side brush and the roller brush assembly, is prone to part of the structure of the temporary obstacle being drawn into the cleaning structure when dealing with some special types of temporary obstacles, such as data cables, causing the sweeping robot to malfunction. Therefore, during the movement process, the cleaning structure is controlled to stop working, and the structural components of the sweeping robot that can fit tightly with the ground, such as the mop structure, are in direct contact with the temporary obstacle, and the temporary obstacle is moved through the movement of the sweeping robot to avoid the occurrence of a malfunction. Avoid the negative impact of temporary obstacles on the sweeping robot. At the same time, determine the area outside the coverage of the normal obstacle closest to the temporary obstacle as the temporary obstacle target area, and control the sweeping robot to move the temporary obstacle into the temporary obstacle target area to avoid the temporary obstacle being continuously blocked by the normal obstacle, which makes it difficult for the user to deal with the temporary obstacle in time. After the temporary obstacle is moved to the temporary obstacle target area, control the sweeping robot to restart the cleaning structure to clean the original cleaning dead zone, and provide the temporary obstacle target area to the user by means of information push, so that the user can know the location of the temporary obstacle in time, so that the user can quickly deal with the temporary obstacle.

[0099] Through this solution, according to the dead zone elimination strategy, the sweeping robot is controlled to move the temporary obstacle, and in this process, the cleaning structure of the sweeping robot is controlled to stop working to avoid the sweeping robot being affected by the temporary obstacle during the movement and causing malfunction. At the same time, the sweeping robot is controlled to move the temporary obstacle to the temporary obstacle target area, and after the sweeping robot completes the cleaning work of the original cleaning dead zone, the temporary obstacle target area is provided to the user, so that the user can know the location of the temporary obstacle in time, so that the user can quickly deal with the temporary obstacle, and effectively avoid the temporary obstacle from continuing to stay in the current area.

[0100] Figure 3 This is a schematic diagram of a multi-dimensional data fusion sweeping robot motion control system provided in one embodiment of the present application, such as Figure 3 As shown, a multi-dimensional data fusion sweeping robot motion control system 300 of this embodiment includes: an obstacle analysis module 301 , a region analysis module 302 , a motion analysis module 303 and a motion control module 304 .

[0101] The obstacle analysis module 301 is used to obtain environmental perception information, analyze the environmental perception information, and determine a temporary obstacle information set and a normal obstacle information set; the area analysis module 302 is used to obtain a robot parameter set, and based on the robot parameter set, determine the cleaning dead zone of the sweeping robot in the obstacle avoidance situation according to the temporary obstacle information set and the normal obstacle information set; the motion analysis module 303 is used to analyze the cleaning dead zone based on the robot parameter set, the environmental perception information and the temporary obstacle information set, and determine the dead zone elimination strategy; the motion control module 304 is used to control the sweeping robot to perform corresponding cleaning motion according to the dead zone elimination strategy, and determine and output the temporary obstacle target area.

[0102] Optionally, the obstacle analysis module 301 is specifically used to: analyze the real-time visual data based on the map data and the real-time position data of the robot to determine the edge features of normal obstacles; analyze the real-time point cloud data based on the normal obstacle edge features to determine the coordinates of the center point of the normal obstacle and the normal obstacle interference radius, and determine the normal obstacle information set according to the normal obstacle edge features and the normal obstacle interference radius; analyze the real-time point cloud data and the real-time visual data to determine the key analysis area of ​​temporary obstacles; adjust the lidar scanning strategy according to the temporary obstacle key analysis area, and determine the temporary obstacle point cloud information according to the adjusted lidar scanning strategy; determine the temporary obstacle type, the temporary obstacle center point coordinates, the temporary obstacle interference radius and the temporary obstacle edge features according to the temporary obstacle point cloud information and the real-time visual data, so as to determine the temporary obstacle information set.

[0103] Optionally, when analyzing the real-time point cloud data and the real-time visual data to determine the key analysis area of ​​the temporary obstacle, the regional analysis module 302 is specifically used to: analyze the real-time visual data to determine the local color features and local edge features of the temporary obstacle; analyze the real-time point cloud data to determine the local geometric features of the temporary obstacle, and perform feature fusion on the local color features of the temporary obstacle, the local edge features of the temporary obstacle and the local geometric features of the temporary obstacle to determine the comprehensive features of the temporary obstacle; perform clustering processing on the comprehensive features of the temporary obstacle according to a preset clustering algorithm to determine the fuzzy area of ​​the temporary obstacle; analyze the fuzzy area of ​​the temporary obstacle to determine the coordinates of the center point of the fuzzy area, and determine the key analysis area of ​​the temporary obstacle according to the preset scanning radius and the coordinates of the center point of the fuzzy area, which is specifically the following formula: ;in, Focus analysis area for the temporary obstacle, is the coordinate of the scanning point to be selected, is the horizontal coordinate of the center point of the fuzzy area, is the ordinate of the center point of the fuzzy area, is the preset scanning radius.

[0104] Optionally, the area analysis module 302 includes a working width and a safe collision distance in the robot parameter set, and analyzes the cleaning dead zone based on the robot parameter set, the environmental perception information and the temporary obstacle information set, and determines the dead zone elimination strategy and the temporary obstacle target area, specifically for: determining the area to be analyzed and the obstacle area according to the normal obstacle edge characteristics and the temporary obstacle edge characteristics; dividing the area to be analyzed into analysis points based on preset analysis point division rules to determine the analysis point coordinate set; determining the obstacle radiation area based on the working width and the safe collision distance according to the normal obstacle edge characteristics, the temporary obstacle edge characteristics and the analysis point coordinate set; determining the cleaning dead zone according to the obstacle area and the obstacle radiation area.

[0105] Optionally, when the area analysis module 302 determines the obstacle radiation area based on the working width and the safe collision distance, according to the normal obstacle edge feature, the temporary obstacle edge feature and the analysis point coordinate set, it is specifically the following formula:

[0106] ;

[0107] in, is the barrier radiation area, is the coordinate of the current analysis point, For the The edge of a temporary obstacle, is the coordinate distance function, For the The edge of a normal obstacle, is the working width, is the safe collision distance; the cleaning dead zone is determined according to the obstacle area and the obstacle radiation area, specifically the following formula: in, To clean up the dead zone, is the barrier radiation area, is the obstacle area.

[0108] Optionally, the motion analysis module 303 is specifically used to: analyze the edge features of the temporary obstacle based on the temporary obstacle type, and determine the coordinates of the movable endpoint of the temporary obstacle; determine the robot's travel path based on the robot's real-time position data and the coordinates of the movable endpoint of the temporary obstacle; determine the movement offset and movement direction based on the robot's travel path, and determine the coordinates of the active endpoint after the movement; determine the safety area based on the robot radius and the normal obstacle interference radius, according to the coordinates of the temporary obstacle center point and the normal obstacle edge features; integrate the robot's travel path, the coordinates of the active endpoint after the movement, and the safety area to determine the dead zone elimination strategy.

[0109] Optionally, the motion analysis module 303 integrates the robot's travel path, the coordinates of the active endpoint after movement, and the safety area to determine the dead zone elimination strategy, specifically the following formula: ;

[0110] Wherein, E is the robot's travel path, is the real-time position data of the robot, For the The coordinates of the temporary obstacle center point of the normal obstacle, is the robot radius, For the The normal obstacle interference radius of the normal obstacle, For the The normal obstacle edge features of the normal obstacles, is the normal obstacle information set, is the coordinate of the active endpoint after the movement, is the temporary obstacle interference radius, is the safe area.

[0111] Optionally, the motion analysis module 303 integrates the robot's travel path, the coordinates of the activity endpoint after movement, and the safety area using the following formula:

[0112] ;

[0113] in, is the robot’s travel path, is the real-time position data of the robot, are the coordinates of the movable endpoints of the temporary obstacle, For the The coordinates of the temporary obstacle center point of the normal obstacle, To make progress for the unit, is the preset maximum step size, is the robot radius, For the The normal obstacle interference radius of the normal obstacle, For the The normal obstacle edge features of the normal obstacles, is the coordinate of the active endpoint after the movement, is the safe area, is the shift offset, is the moving direction, The coordinates of the current analysis point.

[0114] Optionally, the motion control module 304 is specifically used to: control the sweeping robot to move along the robot's travel path to move the temporary obstacle according to the dead zone elimination strategy; when the sweeping robot moves the temporary obstacle, control the cleaning structure of the sweeping robot to stop working, and according to the normal obstacle edge characteristics, determine the area outside the normal obstacle coverage range closest to the temporary obstacle as a temporary obstacle target area, and control the sweeping robot to move the temporary obstacle into the temporary obstacle target area; after the sweeping robot moves the temporary obstacle into the temporary obstacle target area, control the sweeping robot to clean the cleaning dead zone, and record and output the temporary obstacle target area.

[0115] The system of this embodiment can be used to execute the method of any of the above embodiments. The implementation principles and technical effects are similar and will not be described in detail here.

Claims

1. A sweeping robot motion control method based on multi-dimensional data fusion, characterized in that: include: Acquire environmental perception information, analyze the environmental perception information, and determine a temporary obstacle information set and a normal obstacle information set; Acquire a robot parameter set, and determine a cleaning dead zone of the sweeping robot in an obstacle avoidance situation based on the robot parameter set and the temporary obstacle information set and the normal obstacle information set; Analyzing the cleaning dead zone based on the robot parameter set, the environmental perception information and the temporary obstacle information set, and determining a dead zone elimination strategy; According to the dead zone elimination strategy, the cleaning robot is controlled to perform corresponding cleaning movements, and a temporary obstacle target area is determined and output; The environmental perception information includes real-time position data of the robot, the robot parameter set also includes a robot radius, and the cleaning dead zone is analyzed based on the robot parameter set, the environmental perception information and the temporary obstacle information set to determine a dead zone elimination strategy, including: based on the type of temporary obstacles, analyzing the edge features of the temporary obstacles to determine the coordinates of the movable endpoints of the temporary obstacles; The robot's travel path is determined according to the robot's real-time position data and the coordinates of the movable endpoints of the temporary obstacle; the movement offset and movement direction of the sweeping robot relative to the temporary obstacle are determined according to the robot's travel path, and the coordinates of the movable endpoint after the movement are determined accordingly; based on the robot radius and the normal obstacle interference radius, the safety area is determined according to the coordinates of the temporary obstacle center point and the normal obstacle edge features; the robot's travel path, the coordinates of the movable endpoint after the movement and the safety area are integrated to determine the dead zone elimination strategy.

2. The method according to claim 1, characterized in that The environmental perception information includes map data, real-time point cloud data, and real-time visual data. The analyzing the environmental perception information to determine the temporary obstacle information set and the normal obstacle information set includes: Analyzing the real-time visual data based on the map data and the real-time position data of the robot to determine the edge features of normal obstacles; Based on the normal obstacle edge features, the real-time point cloud data is analyzed to determine the normal obstacle center point coordinates and the normal obstacle interference radius, and the normal obstacle information set is determined according to the normal obstacle edge features and the normal obstacle interference radius; Analyze the real-time point cloud data and the real-time visual data to determine a key analysis area for temporary obstacles; According to the temporary obstacle key analysis area, adjusting the laser radar scanning strategy, and determining the temporary obstacle point cloud information according to the adjusted laser radar scanning strategy; According to the temporary obstacle point cloud information and the real-time visual data, the temporary obstacle type, the temporary obstacle center point coordinates, the temporary obstacle interference radius and the temporary obstacle edge features are determined to determine the temporary obstacle information set.

3. The method according to claim 2, characterized in that The analyzing the real-time point cloud data and the real-time visual data to determine a temporary obstacle key analysis area includes: Analyze the real-time visual data to determine local color features and local edge features of the temporary obstacle; Analyze the real-time point cloud data to determine the local geometric features of the temporary obstacle, and perform feature fusion on the local color features of the temporary obstacle, the local edge features of the temporary obstacle, and the local geometric features of the temporary obstacle to determine the comprehensive features of the temporary obstacle; According to a preset clustering algorithm, clustering processing is performed on the comprehensive features of the temporary obstacle to determine the fuzzy area of ​​the temporary obstacle; Analyze the temporary obstacle fuzzy area, determine the coordinates of the center point of the fuzzy area, and determine the temporary obstacle key analysis area according to the preset scanning radius and the coordinates of the center point of the fuzzy area, specifically the following formula: ； in, Focus analysis area for the temporary obstacle, is the coordinate of the scanning point to be selected, is the horizontal coordinate of the center point of the fuzzy area, is the ordinate of the center point of the fuzzy area, is the preset scanning radius.

4. The method according to claim 3, characterized in that: The robot parameter set includes a working width and a safe collision distance, and the determining, based on the robot parameter set and according to the temporary obstacle information set and the normal obstacle information set, of a cleaning dead zone of the sweeping robot in an obstacle avoidance situation includes: Determining the area to be analyzed and the obstacle area according to the normal obstacle edge features and the temporary obstacle edge features; Based on a preset analysis point division rule, the area to be analyzed is divided into analysis points to determine an analysis point coordinate set; Based on the working width and the safe collision distance, determining the obstacle radiation area according to the normal obstacle edge characteristics, the temporary obstacle edge characteristics and the analysis point coordinate set; The cleaning dead zone is determined according to the obstacle area and the obstacle radiation area.

5. The method according to claim 4, characterized in that Based on the working width and the safe collision distance, the obstacle radiation area is determined according to the normal obstacle edge features, the temporary obstacle edge features and the analysis point coordinate set, specifically the following formula: ； in, is the barrier radiation area, is the coordinate of the current analysis point, For the The edge of a temporary obstacle, is the coordinate distance function, For the The edge of a normal obstacle, is the working width, is the safe collision distance; The cleaning dead zone is determined according to the obstacle area and the obstacle radiation area, specifically, the following formula: ； in, To clean up the dead zone, is the barrier radiation area, is the obstacle area.

6. The method according to claim 5, characterized in that The dead zone elimination strategy is determined by integrating the robot's travel path, the coordinates of the active endpoint after the movement, and the safety area, which is specifically the following formula: ； in, is the robot’s travel path, is the real-time position data of the robot, For the The coordinates of the temporary obstacle center point of the normal obstacle, is the robot radius, For the The normal obstacle interference radius of the normal obstacle, For the The normal obstacle edge features of the normal obstacles, is the normal obstacle information set, is the coordinate of the active endpoint after the movement, is the temporary obstacle interference radius, is the safe area.

7. The method according to claim 6, characterized in that The integration of the robot's travel path, the coordinates of the active endpoint after movement, and the safety area is specifically the following formula: ； in, is the robot’s travel path, is the real-time position data of the robot, are the coordinates of the movable endpoints of the temporary obstacle, For the The coordinates of the temporary obstacle center point of the normal obstacle, To make progress for the unit, is the preset maximum step size, is the robot radius, For the The normal obstacle interference radius of the normal obstacle, For the The normal obstacle edge features of the normal obstacles, is the coordinate of the active endpoint after the movement, is the safe area, is the shift offset, is the moving direction, The coordinates of the current analysis point.

8. The method according to claim 7, characterized in that According to the dead zone elimination strategy, controlling the cleaning robot to perform corresponding cleaning motion, determining and outputting a temporary obstacle target area, includes: According to the dead zone elimination strategy, the cleaning robot is controlled to move along the robot's travel path to move the temporary obstacle; When the cleaning robot moves the temporary obstacle, the cleaning structure of the cleaning robot is controlled to stop working, and according to the edge characteristics of the normal obstacle, an area outside the coverage range of the normal obstacle closest to the temporary obstacle is determined as a temporary obstacle target area, and the cleaning robot is controlled to move the temporary obstacle into the temporary obstacle target area; After the cleaning robot moves the temporary obstacle into the temporary obstacle target area, the cleaning robot is controlled to clean the cleaning dead zone, and the temporary obstacle target area is recorded and output.

9. A multi-dimensional data fusion sweeping robot motion control system, characterized in that: The method as claimed in any one of claims 1 to 8 comprises: An obstacle analysis module, used to obtain environmental perception information, analyze the environmental perception information, and determine a temporary obstacle information set and a normal obstacle information set; A region analysis module, used for obtaining a robot parameter set, and determining a cleaning dead zone of the sweeping robot in an obstacle avoidance situation based on the robot parameter set, the temporary obstacle information set and the normal obstacle information set; A motion analysis module, configured to analyze the cleaning dead zone and determine a dead zone elimination strategy based on the robot parameter set, the environmental perception information and the temporary obstacle information set; The motion control module is used to control the cleaning robot to perform corresponding cleaning motion according to the dead zone elimination strategy, and determine and output the temporary obstacle target area.

Images