Obstacle avoidance method and system of double-arm robot

Abstract

The invention discloses an obstacle avoidance method and system for a double-arm robot, relates to the technical field of obstacle avoidance methods for double-arm robots, and aims to determine a three-dimensional avoidance logic according to the moving position of the double-arm robot, the spatial positions of two mechanical arms of the double-arm robot and a plurality of obstacle areas and ensure the accuracy of the three-dimensional avoidance logic. Therefore, the anti-interference logic between the two mechanical arms is determined according to the working paths of the two mechanical arms and the movement ranges of the two mechanical arms; the multi-dimensional moving range of the double-arm robot is determined based on the moving direction of the double-arm robot, the form of the double-arm robot and the moving range of the two mechanical arms, and multiple obstacle avoidance of the double-arm robot relative to each obstacle area is triggered based on the multi-dimensional moving range of the double-arm robot, anti-interference logic and three-dimensional avoidance logic. The intelligence of multiple obstacle avoidance of the double-arm robot relative to each obstacle area is ensured, and the obstacle avoidance accuracy of the double-arm robot relative to each obstacle area is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of obstacle avoidance methods for dual-arm robots, and particularly to an obstacle avoidance method and system for dual-arm robots. Background Art

[0002] With the development of technology, robots are widely used in people's lives and are presented in industrial or commercial scenarios. As a type of dual-arm robot, a dual-arm robot is provided with a mobile base and two robotic arms. The mobile base and the two robotic arms each have a corresponding activity range. In the prior art, when a dual-arm robot detects an obstacle ahead, the dual-arm robot only avoids it based on the activity range of the mobile base of the dual-arm robot, realizing a single-dimensional avoidance logic, without fully considering the activity ranges of the two robotic arms, and unable to ensure the accuracy of the three-dimensional avoidance logic. Summary of the Invention

[0003] The purpose of the present invention is to overcome the deficiencies of the prior art, and the present invention provides an obstacle avoidance method and system for dual-arm robots.

[0004] An embodiment of the present invention provides an obstacle avoidance method for a dual-arm robot, including: determining the actual movement path of the dual-arm robot according to the current position and the target position of the dual-arm robot; determining the current positions of multiple obstacles based on the obstacle detection of the actual movement path of the dual-arm robot, and determining multiple obstacle regions according to the current positions of the multiple obstacles and the shapes of the multiple obstacles; during the movement of the dual-arm robot, determining a three-dimensional avoidance logic according to the movement position of the dual-arm robot, the spatial positions of the two robotic arms of the dual-arm robot, and the multiple obstacle regions; in the two robotic arms of the dual-arm robot, determining an anti-interference logic between the two robotic arms according to the working paths and the activity ranges of the two robotic arms; determining the multi-dimensional activity range of the dual-arm robot based on the movement direction of the dual-arm robot, the shape of the dual-arm robot, and the activity ranges of the two robotic arms, and triggering multiple obstacle avoidances of the dual-arm robot relative to each obstacle region based on the multi-dimensional activity range, the anti-interference logic, and the three-dimensional avoidance logic of the dual-arm robot.

[0005] An embodiment of the present invention provides an obstacle avoidance system for a dual-arm robot. The obstacle avoidance system for the dual-arm robot is applied to the above-mentioned obstacle avoidance method for the dual-arm robot, and the obstacle avoidance system for the dual-arm robot includes: A movement path module, configured to determine the actual movement path of the dual-arm robot according to the current position and the target position of the dual-arm robot; An obstacle region module, configured to determine the current positions of multiple obstacles based on the obstacle detection of the actual movement path of the dual-arm robot, and determine multiple obstacle regions according to the current positions of the multiple obstacles and the shapes of the multiple obstacles; The three-dimensional avoidance module is used to determine the three-dimensional avoidance logic according to the moving position of the dual-arm robot, the spatial positions of the two robotic arms of the dual-arm robot, and multiple obstacle areas during the movement of the dual-arm robot; The anti-interference logic module is used to determine the anti-interference logic between the two robotic arms of the dual-arm robot according to the working paths of the two robotic arms and the moving ranges of the two robotic arms; The multiple obstacle avoidance module is used to determine the multi-dimensional moving range of the dual-arm robot based on the moving direction of the dual-arm robot, the form of the dual-arm robot, and the moving ranges of the two robotic arms, and trigger the multiple obstacle avoidance of the dual-arm robot relative to each obstacle area based on the multi-dimensional moving range of the dual-arm robot, the anti-interference logic, and the three-dimensional avoidance logic.

[0006] Compared with the prior art, the beneficial effects of the present invention are: In the embodiment of the present invention, by the method in the embodiment of the present invention, during the movement of the dual-arm robot, the three-dimensional avoidance logic is determined according to the moving position of the dual-arm robot, the spatial positions of the two robotic arms of the dual-arm robot, and multiple obstacle areas, which accommodates the overall consideration of the moving position of the dual-arm robot, the spatial positions of the two robotic arms of the dual-arm robot, and multiple obstacle areas, and ensures the accuracy of the three-dimensional avoidance logic.

[0007] Therefore, in the two robotic arms of the dual-arm robot, the anti-interference logic between the two robotic arms is determined according to the working paths of the two robotic arms and the moving ranges of the two robotic arms; the multi-dimensional moving range of the dual-arm robot is determined based on the moving direction of the dual-arm robot, the form of the dual-arm robot, and the moving ranges of the two robotic arms, and the multiple obstacle avoidance of the dual-arm robot relative to each obstacle area is triggered based on the multi-dimensional moving range of the dual-arm robot, the anti-interference logic, and the three-dimensional avoidance logic, which accommodates the overall consideration of the multi-dimensional moving range of the dual-arm robot, the anti-interference logic, and the three-dimensional avoidance logic, ensures the intelligence of the multiple obstacle avoidance of the dual-arm robot relative to each obstacle area, and improves the accuracy of the obstacle avoidance of the dual-arm robot relative to each obstacle area. Description of the Drawings

[0008] Figure 1 is a schematic flowchart of the obstacle avoidance method of the dual-arm robot in the embodiment of the present invention; Figure 2 is a schematic flowchart of step S11 in the obstacle avoidance method of the dual-arm robot in the embodiment of the present invention; Figure 3 is a schematic flowchart of step S12 in the obstacle avoidance method of the dual-arm robot in the embodiment of the present invention; Figure 4 is a schematic flowchart of step S13 in the obstacle avoidance method of the dual-arm robot in the embodiment of the present invention; Figure 5 It is a schematic flowchart of step S14 in the obstacle avoidance method of the dual-arm robot in the embodiment of the present invention; Figure 6 It is a schematic flowchart of step S15 in the obstacle avoidance method of the dual-arm robot in the embodiment of the present invention; Figure 7 It is a schematic diagram of the structural composition of the obstacle avoidance system of the dual-arm robot in the embodiment of the present invention. Detailed implementation manners

[0009] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention.

[0010] Please refer to Figures 1 to 7 , an obstacle avoidance method for a dual-arm robot, including: Step S11: Determine the actual movement path of the dual-arm robot according to the current position and the target position of the dual-arm robot; Step S12: Determine the current positions of multiple obstacles based on the obstacle detection of the actual movement path of the dual-arm robot, and determine multiple obstacle regions according to the current positions of the multiple obstacles and the shapes of the multiple obstacles; Step S13: During the movement of the dual-arm robot, determine the three-dimensional avoidance logic according to the movement position of the dual-arm robot, the spatial positions of the two robotic arms of the dual-arm robot, and multiple obstacle regions; Step S14: In the two robotic arms of the dual-arm robot, determine the anti-interference logic between the two robotic arms according to the working paths of the two robotic arms and the movement ranges of the two robotic arms; Step S15: Determine the multi-dimensional movement range of the dual-arm robot based on the movement direction of the dual-arm robot, the shape of the dual-arm robot, and the movement ranges of the two robotic arms, and trigger multiple obstacle avoidances of the dual-arm robot relative to each obstacle region based on the multi-dimensional movement range, the anti-interference logic, and the three-dimensional avoidance logic of the dual-arm robot; Refer to Figure 2 , in step S11, determine the actual movement path of the dual-arm robot according to the current position and the target position of the dual-arm robot; In the specific implementation process of the present invention, the specific steps are as follows: S111: Collect the current position and the target position of the dual-arm robot, determine the movement space of the dual-arm robot according to the current position and the target position of the dual-arm robot, and determine multiple alternative movement paths according to the movement space and the previous movement path of the dual-arm robot; S112: Determine 3 better movement paths according to the preliminary screening of the detections of the multiple alternative movement paths, and the matching coefficients of the 3 better movement paths with the dual-arm robot meet a preset matching coefficient threshold; S113: Determine the moving distances of each optimal moving path based on the detection of 3 optimal moving paths, and determine the actual moving path of the dual-arm robot according to the moving distances of each optimal moving path, the form of the dual-arm robot, and the power of the dual-arm robot.

[0011] In the embodiments of the present application, collect the current position and target position of the dual-arm robot, and obtain the accurate position information of the dual-arm robot through sensors or positioning systems (such as GPS, lidar, inertial navigation unit, etc.); the current position is the current coordinate or position description of the dual-arm robot, and the target position is the end coordinate or position description that the dual-arm robot needs to reach; at this time, an actual example: assume that the dual-arm robot is performing a handling task in a warehouse, its current position is the entrance of the warehouse (coordinates: X1, Y1), and the target position is a shelf in the warehouse (coordinates: X2, Y2); through GPS or lidar positioning, the dual-arm robot can accurately know its current position and target position.

[0012] According to the current position and the target position, combined with the environmental map or real-time sensor data, determine the space range occupied by the dual-arm robot during movement. This space range should take into account the size, form, and movement trajectory of the dual-arm robot; at this time, in the warehouse environment, the movement space of the dual-arm robot is a rectangular area on a two-dimensional plane. This area starts from the current position and ends at the target position, and takes into account the width and length of the dual-arm robot, as well as the turning radius; for example, if the width of the dual-arm robot is W, the length is L, and the turning radius is R, then the movement space is a rectangle with a width of W + 2*R (considering turning) and a length equal to the shortest distance from the current position to the target position.

[0013] Use path planning algorithms to generate multiple moving paths within the determined movement space. These paths should consider the limitations such as the form, size, movement speed, and turning ability of the dual-arm robot; at the same time, also combine the previous moving path data of the dual-arm robot, and use machine learning or heuristic search algorithms to optimize path selection; at this time, in the warehouse, there are multiple paths for the dual-arm robot to reach the target position from the current position.

[0014] For example, it chooses to move straight forward directly (if there are no obstacles), or bypass a certain obstacle; based on its previous moving experience, the dual-arm robot knows which paths are feasible and which paths encounter obstacles; assume that the dual-arm robot has successfully walked two paths in the past: one is to move straight forward (Path A), and the other is to bypass an obstacle (Path B); in this task, the dual-arm robot also considers a new path (Path C), which combines parts of Path A and Path B to find a better solution; therefore, the dual-arm robot will generate these three alternative paths and be ready for further evaluation and selection.

[0015] Further, three preferred movement paths are determined based on the preliminary screening of the detection of multiple alternative movement paths, and the matching coefficients of the three preferred movement paths with the dual-arm robot meet a preset matching coefficient threshold.

[0016] At this time, the dual-arm robot will detect the multiple alternative movement paths generated previously, which usually includes the detection of potential obstacles on the path, the calculation of the path length, the evaluation of the path curvature, etc.; the purpose of the detection is to evaluate the feasibility, safety, and efficiency of each path; at this time, the dual-arm robot will use sensors (such as lidar, cameras) to detect obstacles on the path in real time, and at the same time combine the environmental map and path planning algorithm to calculate the length and curvature of the path; in addition, the dual-arm robot also considers other factors, such as the lighting conditions on the path, the ground material, etc., which will affect the movement speed and stability of the dual-arm robot.

[0017] After detecting all the alternative paths, the dual-arm robot will conduct a preliminary screening of the paths according to the preset screening criteria; the screening criteria include the safety of the path (such as avoiding collision with obstacles), efficiency (such as path length and required time), and the adaptability of the dual-arm robot (such as considering the shape, size, and movement ability of the dual-arm robot); at this time, the dual-arm robot will score each path, and the higher the score, the better the path; the scoring is based on the weighted sum of multiple indicators, such as the reciprocal of the path length (the shorter the higher the score), the reciprocal of the number of obstacles on the path (the fewer the higher the score), and the reciprocal of the path curvature (the smoother the higher the score); then, the dual-arm robot will select several paths with the highest scores as candidates.

[0018] After the preliminary screening, the dual-arm robot will select three preferred movement paths from the candidate paths. These paths should not only perform well in terms of safety, efficiency, and adaptability, but also be highly matched with the current state and task requirements of the dual-arm robot; in order to quantify this matching degree, the dual-arm robot will calculate the matching coefficient of each path; at this time, the matching coefficient is a comprehensive score, considering the matching degree of the path with multiple factors such as the shape, size, movement speed, battery power, and task priority of the dual-arm robot; for example, if a path requires the dual-arm robot to turn frequently, and the turning ability of the dual-arm robot is limited, then the matching coefficient of this path will be lower; on the contrary, if a path is straight and short, and the dual-arm robot has sufficient battery power, then the matching coefficient of this path will be higher; in order to ensure that the selected path is feasible and efficient, the dual-arm robot will set a preset matching coefficient threshold; only the paths with a matching coefficient higher than this threshold will be selected as the preferred movement paths.

[0019] Specifically, the dual-arm robot has generated multiple alternative movement paths; in step S112, the dual-arm robot starts to detect and screen these paths; the dual-arm robot uses lidar and cameras to detect obstacles on the paths in real time, and combines the environmental map to calculate the length and curvature of each path; for example, the dual-arm robot finds that there is a large obstacle on path A that needs to be bypassed, while path B is relatively smooth and obstacle-free although it is longer; based on the detection results, the dual-arm robot scores each path; path A scores lower because it needs to bypass the obstacle; path B scores medium although it is long but safe and smooth; and path C (a newly discovered short and straight path) scores the highest because it is short and obstacle-free; after the preliminary screening, the dual-arm robot selects the three paths with the highest scores (assumed to be path B, C, and another alternative path D) as candidates; then, the dual-arm robot calculates the matching coefficient of each path, considering factors such as the shape, size, movement speed, battery level, and task priority of the dual-arm robot; finally, the dual-arm robot determines paths C, B, and D as the better movement paths because their matching coefficients are all higher than the preset threshold.

[0020] Therefore, based on the detection of the three better movement paths, the movement distances of each better movement path are determined, and according to the movement distances of each better movement path, the shape of the dual-arm robot, and the battery level of the dual-arm robot, the actual movement path of the dual-arm robot is determined, which takes into account the overall consideration of the movement distances of each better movement path, the shape of the dual-arm robot, and the battery level of the dual-arm robot, ensuring the accuracy of the actual movement path of the dual-arm robot.

[0021] At this time, the dual-arm robot will calculate the actual movement distance of each path based on the three better movement paths determined before. This is usually an accurate measurement of the path length, considering the straight and curved parts of the path, as well as obstacle bypassing; at this time, the dual-arm robot uses the distance calculation function in the path planning algorithm, combines the environmental map and sensor data, to accurately measure the length of each path; for the curved part and obstacle bypassing.

[0022] When determining the actual movement path, the dual-arm robot needs to comprehensively consider its own form and power; form factors include the size, weight, and range of motion of the robotic arms of the dual-arm robot, and these factors will affect the movement efficiency and stability of the dual-arm robot on different paths; the power factor involves the remaining power and estimated power consumption of the dual-arm robot, which will affect the endurance and reliability of the dual-arm robot during task execution; at this time, the dual-arm robot will assign an additional weight or score to each path according to the form and power factors; for example, if a path requires the dual-arm robot to move or turn frequently, and the form or power of the dual-arm robot is not suitable for this high-intensity movement, then the weight of this path will be reduced; on the contrary, if a path is relatively flat and does not require much movement, then the weight of this path will be increased.

[0023] After comprehensively considering the movement distance, form, and power factors, the dual-arm robot will select an optimal path as the actual movement path, and this path should be the best balance in terms of safety, efficiency, and adaptability, and can meet the current state and task requirements of the dual-arm robot; at this time, the dual-arm robot will adopt decision-making methods such as weighted sum or sorting to determine the optimal path; the weighted sum method involves adding the weights of factors such as movement distance, form, and power to obtain a comprehensive score; then, the dual-arm robot will select the path with the highest score as the actual movement path; the sorting method involves sorting the paths according to one or more key factors (such as movement distance or form adaptability), and then selecting the path ranked at the top.

[0024] Reference Figure 3 , in step S12, based on the obstacle detection of the actual movement path of the dual-arm robot, the current positions of multiple obstacles are determined, and multiple obstacle regions are determined according to the current positions of the multiple obstacles and the forms of the multiple obstacles; In the specific implementation process of the present invention, the specific steps are as follows: S121: Collect the actual movement path of the dual-arm robot, and mark multiple obstacles on the actual movement path according to the detection of the actual movement path of the dual-arm robot, and determine the current positions of the multiple obstacles; the current positions of the multiple obstacles include the horizontal azimuth and height positions of the multiple obstacles; S122: The dual-arm robot is equipped with a radar, and according to the detection of each obstacle by the radar, the corresponding point cloud set is collected, and the form and type of the obstacle are determined according to the recognition of each point cloud set; S123: Determine the activity influence range corresponding to the obstacle based on the form and type of the obstacle, and determine multiple obstacle regions according to the horizontal azimuth, height position of the multiple obstacles, and the activity influence range corresponding to the obstacle; In an embodiment of the present application, the dual-arm robot utilizes its built-in navigation system and sensors (such as GPS, inertial navigation unit, lidar, etc.) to collect its moving path in the actual environment in real time. These sensors provide key information such as the position, speed, and direction of the dual-arm robot, thereby allowing the dual-arm robot to construct and update the map of its current moving path. At this time, the dual-arm robot uses SLAM (Simultaneous Localization and Mapping) technology to construct the environmental map in real time and update the map according to its own position. In addition, the dual-arm robot also uses auxiliary means such as vision sensors or RFID tags to improve the positioning accuracy.

[0025] While collecting the moving path, the dual-arm robot uses its sensors (such as lidar, camera, etc.) to detect obstacles on the path. These sensors emit signals (such as laser beams or light) into the surrounding environment and receive the reflected signals to identify obstacles. At this time, the lidar calculates the distance to the obstacle by emitting laser beams and measuring the reflection time. The camera detects obstacles by capturing environmental images and using image processing algorithms. In addition, the dual-arm robot also uses sonar sensors to detect obstacles at close range.

[0026] Once an obstacle is detected, the dual-arm robot marks it on the environmental map and determines the current position of the obstacle, which includes the horizontal orientation (such as X, Y coordinates) and height position (such as Z coordinate) of the obstacle, so that the dual-arm robot can accurately avoid it in subsequent planning. At this time, the dual-arm robot calculates the position of the obstacle based on the data provided by the sensors. For the lidar, the dual-arm robot uses triangulation or time-difference measurement methods to calculate the distance and height of the obstacle. For the camera, the dual-arm robot uses image processing algorithms (such as edge detection, feature matching, etc.) to identify the contour and position of the obstacle.

[0027] Specifically, assume that the dual-arm robot is performing a handling task in a warehouse, and its actual moving path is from shelf A to shelf B. During the movement, the dual-arm robot needs to avoid obstacles on the path. The dual-arm robot uses its built-in navigation system and lidar to collect its moving path in real time. During the movement, the dual-arm robot continuously updates its position and the environmental map. When the dual-arm robot approaches the passage between shelf A and shelf B, its lidar detects that there is a large cargo box (obstacle 1) blocking the way in the passage. At the same time, the camera also captures an image of the cargo box. The dual-arm robot marks the cargo box on the environmental map and calculates its current position. Assume that the horizontal orientation of the cargo box is (X1, Y1) and the height position is Z1, and this information will be used for subsequent path planning and obstacle avoidance operations.

[0028] Furthermore, the dual-arm robot is equipped with a radar. According to the detection of each obstacle by the radar, a corresponding point cloud set is collected, and based on the recognition of each point cloud set, the shape and type of the obstacle are determined.

[0029] At this time, the radar (usually a lidar) equipped on the dual-arm robot will actively emit laser beams into the surrounding environment and receive the signals reflected back from the surface of the obstacles by these laser beams. The radar calculates the distance and azimuth of the obstacles by measuring the time difference between the emission and reception of the laser beams and the deflection angle of the laser beams. At this time, the lidar continuously emits laser beams into the surrounding environment in a rotating or scanning manner to form a continuous scanning plane. When the laser beam encounters an obstacle, it will be reflected back and received by the radar. The processor inside the radar will calculate the three-dimensional coordinates of the obstacle based on the received signal.

[0030] While detecting obstacles, the radar will collect a large amount of three-dimensional point data, which is called a point cloud set. Each point in the point cloud set represents a position on the surface of the obstacle. Through the set of these points, the three-dimensional shape of the obstacle is constructed. At this time, the lidar will continuously emit and receive laser beams. Each time a reflected signal is received, the coordinates of a three-dimensional point will be calculated. With the continuous scanning of the radar, a point cloud set containing a large number of points will be generated, and these point cloud data are stored in the memory of the dual-arm robot for subsequent processing and analysis.

[0031] The dual-arm robot will process and analyze the collected point cloud set to identify the shape and type of the obstacle. This usually involves operations such as filtering, segmentation, feature extraction, and classification of the point cloud data. The dual-arm robot uses point cloud processing algorithms (such as algorithms in the PCL library) to process the point cloud data. First, noise points and outliers are removed through the filtering operation. Then, the segmentation algorithm is used to divide the point cloud data into different parts, with each part corresponding to an obstacle. Next, the features of each part (such as shape, size, surface texture, etc.) are extracted. Finally, classification algorithms (such as support vector machines, random forests, etc.) are used to classify the obstacles according to the features to determine their types and shapes.

[0032] Therefore, based on the shape and type of the obstacle, the corresponding activity influence range of the obstacle is determined, and based on the horizontal azimuth, height position of multiple obstacles, and the activity influence range corresponding to the obstacles, multiple obstacle areas are determined, taking into account the overall consideration of the horizontal azimuth, height position of multiple obstacles, and the activity influence range corresponding to the obstacles, ensuring the accuracy of multiple obstacle areas.

[0033] At this time, the dual-arm robot will evaluate the active influence range of each obstacle according to the shape and type of the obstacles detected and recognized by the radar before; the active influence range refers to the area around the obstacle that interferes with its movement or operation, and the size and shape of this range depend on the specific characteristics of the obstacle, such as size, shape, stability, and whether it is moving, etc.; at this time, there is a predefined database of obstacle influence ranges inside the dual-arm robot. According to the type and shape of the obstacle, the dual-arm robot looks up the corresponding active influence range template from the database; if there is no directly matching template in the database, the dual-arm robot will use algorithms based on rules or machine learning to dynamically calculate the active influence range.

[0034] The dual-arm robot will use its navigation system or sensors to accurately determine the horizontal orientation (X, Y coordinates) and height position (Z coordinate) of each obstacle, and this information is the basis for subsequent division of the obstacle area; at this time, the dual-arm robot will use GPS, inertial navigation unit, lidar or other types of sensors to obtain the position information of the obstacle; for the determination of the height position, sensors that can measure distance, such as lidar or ultrasonic sensors, are particularly useful.

[0035] The dual-arm robot will divide the corresponding obstacle areas in the environmental map according to the active influence range and position information of each obstacle, and these areas will be used as references for the dual-arm robot in subsequent path planning and obstacle avoidance operations; at this time, the dual-arm robot will mark the active influence range of each obstacle in its internal environmental map in the form of polygons or other shapes, and these marks are dynamically updated to reflect changes in the position or shape of the obstacle; when planning a path, the dual-arm robot will avoid these obstacle areas to ensure safe and efficient task completion.

[0036] At this time, assume that the dual-arm robot is performing a handling task in a complex warehouse environment, and there are multiple different types of obstacles in the warehouse, such as shelves, large cargo boxes, and mechanical equipment, etc.; the dual-arm robot first detects and recognizes the obstacles in the warehouse through the radar; for the shelves, the dual-arm robot knows that they are usually stationary, so the active influence range is limited to the occupied space of the shelves themselves; for the large cargo boxes, the dual-arm robot will evaluate their stability and the swaying range caused by movement, so as to determine a slightly larger active influence range; for the mechanical equipment, the dual-arm robot will consider its operating radius and the movement range generated, so as to determine a wider active influence range.

[0037] The dual-arm robot uses its navigation system to accurately determine the horizontal orientation and height position of each obstacle; for example, the coordinates of a shelf are (X1, Y1, Z1), the coordinates of a large cargo box are (X2, Y2, Z2), and the coordinates of a mechanical device are (X3, Y3, Z3); according to the active influence range and position information of the obstacles, the dual-arm robot divides the corresponding obstacle areas in its internal environment map, and these areas are marked with polygons or other shapes and are dynamically updated to reflect changes in the position or form of the obstacles; when planning the subsequent path, the dual-arm robot will avoid these obstacle areas to ensure the safe execution of tasks; through the above steps, the dual-arm robot can determine its active influence range based on the form and type of the obstacles and divide the obstacle areas according to the position information, which provides important support for the path planning and obstacle avoidance operations of the dual-arm robot in a complex environment.

[0038] In an embodiment of the present application, an active influence range matching table is collected, and the active influence range matching table is shown in Table 1: Table 1 Active Influence Range Matching Table Obstacle type Activity influence range Shelving Rectangular area with the same size as the shelving Large cargo box Rectangular area expanded by a certain distance based on the cargo box size Mechanical equipment Circular or elliptical area centered on the equipment, considering the operation radius Suppose the dual-arm robot detects a large cargo box with dimensions of 2m × 3m × 1.5m (length × width × height); according to the active influence range matching table, the dual-arm robot will find a rectangular area template that expands a certain distance outward based on the size of the cargo box. For example, if it expands 0.5m outward, the active influence range is a rectangular area of 2.5m × 3.5m; combined with the actual position of the cargo box (such as (X, Y, Z) coordinates), the dual-arm robot divides the corresponding obstacle area in the environmental map.

[0039] Reference Figure 4 , in step S13, during the movement of the dual-arm robot, a three-dimensional avoidance logic is determined according to the movement position of the dual-arm robot, the spatial positions of the two robotic arms of the dual-arm robot, and multiple obstacle areas; In the specific implementation process of the present invention, the specific steps are as follows: S131: Real-time monitor the movement of the dual-arm robot, collect the movement position of the dual-arm robot, and determine multiple obstacle areas around the dual-arm robot based on the detection of the periphery of the movement position of the dual-arm robot. At this time, based on the detection of the dual-arm robot and multiple obstacle areas around the dual-arm robot, determine the active range of the dual-arm robot during movement; S132: In the dual-arm robot, construct a corresponding dual-arm robot coordinate system based on the dual-arm robot, and determine the spatial positions of the two robotic arms of the dual-arm robot according to the two robotic arms of the dual-arm robot and the dual-arm robot coordinate system; S133: Determine the active dimensions of the two robotic arms of the dual-arm robot based on the traversal of the two robotic arms, and determine the active ranges of the two robotic arms of the dual-arm robot according to the active dimensions of the two robotic arms and the relative distance between the two robotic arms, and determine the three-dimensional avoidance logic according to the active range during the movement of the dual-arm robot and the active ranges of the two robotic arms of the dual-arm robot.

[0040] In an embodiment of the present application, the movement of the dual-arm robot is monitored in real time, the movement position of the dual-arm robot is collected, and a plurality of obstacle areas around the dual-arm robot are determined based on the detection of the periphery of the movement position of the dual-arm robot. At this time, based on the detection of the dual-arm robot and the plurality of obstacle areas around the dual-arm robot, the active range during the movement of the dual-arm robot is determined, which incorporates the overall consideration of the detection of the dual-arm robot and the plurality of obstacle areas around the dual-arm robot, ensuring the accuracy of the active range during the movement of the dual-arm robot.

[0041] At this time, the dual-arm robot is equipped with a variety of sensors (such as encoders, gyroscopes, odometers, etc.), and these sensors can monitor the movement state of the dual-arm robot in real time, including information such as position, speed, and acceleration. In addition, the dual-arm robot also uses an external positioning system (such as GPS, UWB positioning system, etc.) to obtain more accurate global position information. At this time, the sensor data is continuously collected and transmitted to the control system of the dual-arm robot. The control system processes these data through algorithms to update the position information of the dual-arm robot in real time. For the external positioning system, the dual-arm robot needs to receive the positioning signal through the communication module and convert it into the position coordinates of the dual-arm robot itself.

[0042] Based on the real-time monitoring, the control system of the dual-arm robot will regularly or as needed collect the current position information of the dual-arm robot, and this information is usually represented in coordinate form, including the X, Y, Z coordinates (and rotation angles) of the dual-arm robot in a two-dimensional plane or three-dimensional space. At this time, the control system will calculate the current position coordinates of the dual-arm robot according to the sensor data or the signal of the external positioning system and store it in the memory for subsequent use.

[0043] During the movement of the dual-arm robot, sensors such as radar, cameras, and lidar are used to scan and detect the surrounding environment. These sensors can identify the obstacles around the dual-arm robot and mark the corresponding obstacle areas in the environmental map of the dual-arm robot according to the shape, type, and position information of the obstacles. At this time, the environmental data collected by the sensors is transmitted to the perception system of the dual-arm robot. The perception system processes these data through algorithms to identify the obstacles and draw the obstacle areas in the environmental map. These areas are usually represented in polygons, circles, or other shapes to reflect the actual occupied space of the obstacles.

[0044] After determining the current position of the dual-arm robot and the surrounding obstacle areas, the control system of the dual-arm robot calculates the movable range of the dual-arm robot during movement based on this information; the movable range refers to the area where the dual-arm robot can move safely without colliding with any obstacles; at this time, the control system uses a path planning algorithm, combines the current position of the dual-arm robot, the obstacle area, and a preset safety distance, and calculates one or more feasible paths from the current position to the target position. The area covered by these paths is the movable range of the dual-arm robot during movement; the control system also continuously updates this range to reflect the changes in the position of the dual-arm robot and the dynamic changes in the surrounding environment.

[0045] At this time, assume that the dual-arm robot is performing a handling task in an automated warehouse; there are multiple shelves, mechanical equipment, and transportation channels in the warehouse; the dual-arm robot uses sensors such as built-in encoders and gyroscopes to monitor its movement state in real time, including information such as position, speed, and acceleration; at the same time, the dual-arm robot also uses a UWB positioning system to obtain its global position coordinates in the warehouse; when the dual-arm robot moves near a certain shelf, the control system collects its current position coordinates as (X, Y, Z), where X and Y represent the position of the dual-arm robot in the two-dimensional plane, and Z represents the height of the dual-arm robot.

[0046] The dual-arm robot uses lidar to scan the surrounding environment and identify obstacles such as nearby shelves and mechanical equipment. These obstacles are marked as polygon areas in the environmental map of the dual-arm robot to reflect their actual occupied space; based on the current position of the dual-arm robot and the obstacle area information, the control system calculates an optimal path from the current position to the shelf. This path avoids all obstacle areas and ensures that the dual-arm robot will not collide with any obstacles during movement; the control system also continuously updates this path and the movable range to reflect the changes in the position of the dual-arm robot and the dynamic changes in the surrounding environment; through the above steps, the dual-arm robot can monitor its movement state in real time and determine its movable range during movement based on the surrounding environment information, which provides important support for the autonomous navigation and obstacle avoidance of the dual-arm robot in a complex environment.

[0047] Furthermore, in the dual-arm robot, a corresponding dual-arm robot coordinate system is constructed based on the dual-arm robot, and the spatial positions of the two robotic arms of the dual-arm robot are determined according to the two robotic arms of the dual-arm robot and the dual-arm robot coordinate system, taking into account the overall consideration of the two robotic arms of the dual-arm robot and the dual-arm robot coordinate system, and ensuring the accuracy of the spatial positions of the two robotic arms of the dual-arm robot.

[0048] At this time, in a dual-arm robot, it is first necessary to construct a coordinate system based on the dual-arm robot itself, that is, the dual-arm robot coordinate system. This coordinate system usually selects a certain fixed point of the dual-arm robot (such as the center of gravity, the center of the base, etc.) as the origin, and defines three mutually perpendicular coordinate axes (X-axis, Y-axis, Z-axis) to describe the position and orientation of the dual-arm robot in three-dimensional space. At this time, the construction of the dual-arm robot coordinate system is usually completed in the design stage of the dual-arm robot and is implemented through programming in the control system of the dual-arm robot. The selection and definition of the coordinate system need to consider the structural characteristics, kinematic model, and subsequent task requirements of the dual-arm robot.

[0049] After constructing the dual-arm robot coordinate system, it is necessary to determine the spatial positions of the two manipulators of the dual-arm robot in this coordinate system. This is usually achieved by using sensors (such as encoders, gyroscopes, etc.) installed on the manipulators to measure the angles or displacements of the joints of the manipulators in real time, and then using the kinematic model of the dual-arm robot to convert these measured values into the three-dimensional coordinates of the end effector of the manipulator in the dual-arm robot coordinate system. At this time, the determination of the spatial position of the manipulator involves the forward kinematics calculation of the dual-arm robot. Forward kinematics refers to calculating the position of the end effector of the manipulator in the dual-arm robot coordinate system based on the angles or displacements of the joints of the manipulator, which is usually achieved through matrix transformation, involving the calculation of rotation matrices and translation vectors.

[0050] Therefore, the active dimensions of the two manipulators of the dual-arm robot are determined based on the traversal of the two manipulators of the dual-arm robot, and the active ranges of the two manipulators of the dual-arm robot are determined according to the active dimensions of the two manipulators and the relative distance between the two manipulators. Then, the three-dimensional avoidance logic is determined based on the active range during the movement of the dual-arm robot and the active ranges of the two manipulators of the dual-arm robot. This takes into account the overall active range during the movement of the dual-arm robot and the active ranges of the two manipulators of the dual-arm robot, ensuring the accuracy of the three-dimensional avoidance logic. At the same time, it takes into account the overall movement position of the dual-arm robot, the spatial positions of the two manipulators of the dual-arm robot, and multiple obstacle areas, ensuring the accuracy of the three-dimensional avoidance logic.

[0051] At this time, the two robotic arms of the dual-arm robot each have a certain range of motion and movement capabilities, which are achieved through the movement of the various joints of the robotic arm (such as rotational joints, translational joints, etc.); the degrees of freedom refer to the directions in which the robotic arm can move freely, which is usually related to the structural design and joint configuration of the robotic arm; for example, a robotic arm with three rotational joints performs rotational movements in three different planes; at this time, determining the degrees of freedom of the robotic arm usually involves the analysis of the kinematic model of the robotic arm; by analyzing the joint configuration and connection method of the robotic arm, determine the directions in which the robotic arm can move, as well as the range and limitations of these movements.

[0052] After determining the degrees of freedom of the robotic arm, it is necessary to further determine the range of motion of the robotic arm in these degrees of freedom; the range of motion refers to the set of all positions that the end effector of the robotic arm can reach, which is usually affected by the robotic arm structure, joint limitations, and the external environment (such as obstacles); at this time, determining the range of motion of the robotic arm usually involves the simulation and analysis of the kinematic model of the robotic arm; by simulating the movement state of the robotic arm at different joint angles, calculate all the positions that the end effector of the robotic arm can reach, and draw a map of the range of motion of the robotic arm.

[0053] After determining the movable range of the dual-arm robot and the range of motion of the two robotic arms, it is necessary to formulate a three-dimensional avoidance logic to ensure that the dual-arm robot does not collide with obstacles or other robotic arms when performing tasks; the three-dimensional avoidance logic usually involves the planning, monitoring, and adjustment of the movement trajectories of the dual-arm robot and the robotic arms; at this time, determining the three-dimensional avoidance logic usually involves the application of path planning algorithms, collision detection algorithms, and real-time control strategies; the path planning algorithm is used to plan the movement trajectories of the dual-arm robot and the robotic arms to ensure that they can reach the target position safely; the collision detection algorithm is used to monitor the distance between the dual-arm robot and the robotic arms and obstacles or other robotic arms in real time, and make adjustments when necessary to avoid collisions; the real-time control strategy is used to quickly respond to and adjust the movement of the dual-arm robot and the robotic arms according to the collision detection results.

[0054] Specifically, for a dual-arm robot, the robotic arm 1 is responsible for grasping parts, and the robotic arm 2 is responsible for installing the parts at the designated position; the movable range of the dual-arm robot has been determined through previous steps; both the robotic arm 1 and the robotic arm 2 have three rotational joints, so they both perform rotational movements on three different planes, and these rotational movements constitute the movement dimensions of the robotic arms; by simulating and analyzing the kinematic models of the robotic arm 1 and the robotic arm 2, their movement ranges at different joint angles are determined; for example, the end effector of the robotic arm 1 can perform rotational and translational movements within a certain radius centered on the base, while the movement range of the robotic arm 2 is affected by its structure and joint limitations, and these movement ranges form the reachable areas of the robotic arms in three-dimensional space.

[0055] When formulating the three-dimensional avoidance logic, first, the motion trajectory planning of the dual-arm robot and the robotic arms needs to be considered; for example, when the robotic arm 1 needs to grasp a part, it will move along a planned path to the position where the part is located; at the same time, the robotic arm 2 will stay at a safe position within its movement range to avoid colliding with the robotic arm 1; during the movement of the dual-arm robot, the collision detection algorithm will continuously monitor the distances between the dual-arm robot, the robotic arms and the obstacles, and make adjustments when necessary to avoid collisions; for example, if the robotic arm 1 detects that the distance to an obstacle during movement is less than the preset safe distance, it will immediately stop moving and make adjustments to ensure that it can continue to perform the task safely; through the above steps, the dual-arm robot can determine the movement dimensions and movement ranges of its two robotic arms, and formulate the corresponding three-dimensional avoidance logic, which provides important support for the precise control and safe obstacle avoidance of the dual-arm robot when performing complex tasks.

[0056] Reference Figure 5 , in step S14, among the two robotic arms of the dual-arm robot, the anti-interference logic between the two robotic arms is determined according to the working paths and movement ranges of the two robotic arms; In the specific implementation process of the present invention, the specific steps are as follows: S141: Determine the movement ranges of the two robotic arms according to the detection of the two robotic arms of the dual-arm robot. At this time, further optimize the movement ranges of the two robotic arms to avoid overlapping areas between the movement ranges of the two robotic arms; S142: Continuously monitor the two robotic arms of the dual-arm robot, collect the work tasks of the dual-arm robot, and determine the sub-work tasks corresponding to the two robotic arms according to the analysis of the work tasks. The sub-work tasks corresponding to the two robotic arms are the first sub-work task and the second sub-work task respectively; S143: Determine the working paths of the two robotic arms based on the first sub-work task, the second sub-work task, and the two robotic arms of the dual-arm robot. Determine multiple anti-interference warning points based on the working paths of the two robotic arms and their movement ranges. Determine the anti-interference logic between the two robotic arms based on the multiple anti-interference warning points and the training of the two robotic arms. This anti-interference logic ensures that the two robotic arms do not come into contact with each other.

[0057] In an embodiment of the present application, determine the movement ranges of the two robotic arms of the dual-arm robot through detection. At this time, further optimize the movement ranges of the two robotic arms to avoid overlapping areas between the movement ranges of the two robotic arms.

[0058] At this time, conduct a detailed kinematic analysis of the two robotic arms of the dual-arm robot; by considering the joint configurations of the robotic arms, joint limitations (such as rotation angles, translation distances, etc.), and external environmental factors (such as obstacle positions, workspace limitations, etc.), determine the set of all positions that each robotic arm can reach in three-dimensional space, that is, its movement range. This usually requires the use of forward kinematics and inverse kinematics knowledge in dual-arm robotics; at this time, in actual operation, model the robotic arms through simulation software and simulate their movement states at different joint angles to calculate the movement range; in addition, also use sensors installed on the robotic arms (such as encoders, gyroscopes, etc.) to collect joint angle data in real time and calculate the current position and movement range of the robotic arms through algorithms.

[0059] After determining the movement ranges of the two robotic arms, further optimization is required to avoid overlapping or conflicting areas between them. This is achieved by adjusting the joint limitations of the robotic arms, changing the base position or direction of the robotic arms, re-planning the workspace layout, etc.; the goal of optimization is to ensure that the two robotic arms can work independently and efficiently within their respective movement ranges without interfering with each other; at this time, the methods for optimizing the movement ranges vary depending on the specific situation; for example, if there is an overlap in the movement ranges of the two robotic arms, consider adjusting the base position or direction of one of the robotic arms to reduce the overlapping area; in addition, also use advanced algorithms (such as genetic algorithms, particle swarm algorithms, etc.) to automatically optimize the movement ranges of the robotic arms to find the optimal workspace layout.

[0060] Furthermore, monitor the two robotic arms of the dual-arm robot in real time, collect the work tasks of the dual-arm robot, and determine the sub-work tasks corresponding to the two robotic arms according to the analysis of the work tasks. The sub-work tasks corresponding to the two robotic arms are the first sub-work task and the second sub-work task respectively.

[0061] At this time, continuous monitoring is carried out on the two robotic arms of the dual-arm robot; the monitoring content includes kinematic parameters such as the position, speed, acceleration, and joint angles of the robotic arms, as well as the data collected by the installed sensors (such as force sensors, vision sensors, etc.); the purpose of real-time monitoring is to grasp the current state of the robotic arms in real time and provide accurate information for subsequent task allocation and path planning. At this time, in actual operation, sensors (such as encoders, gyroscopes, accelerometers, etc.) are installed on the robotic arms to collect kinematic parameters in real time; in addition, machine vision technologies (such as cameras, lidar, etc.) are also used to monitor changes in the environment around the robotic arms. These sensors and data acquisition devices are usually connected to the control system of the dual-arm robot to enable real-time transmission and processing of data.

[0062] Receive the work tasks of the dual-arm robot from an external system or user. These tasks are predefined, periodic tasks (such as assembly tasks on a production line), and also tasks dynamically generated according to real-time requirements (such as picking and packing specific products according to order requirements); the tasks are usually transmitted to the control system of the dual-arm robot in the form of instructions or data packets; at this time, task acquisition is achieved through various methods, such as receiving task instructions from a remote server via wireless communication (such as Wi-Fi, Bluetooth, etc.), or data transmission with a host connected to the control system through a wired connection (such as Ethernet); in addition, natural language processing (NLP) technologies are also used to parse the user's voice instructions, or task inputs from the user are received through a graphical user interface (GUI).

[0063] After receiving the work tasks, it is necessary to parse the tasks to determine the sub-work tasks that each robotic arm needs to complete. This step performs operations such as decomposing, sorting, and prioritizing the tasks; the purpose of parsing is to decompose complex tasks into a series of simple actions or steps that can be executed by the robotic arms and ensure that these actions or steps can be efficiently completed without interfering with each other; at this time, task parsing is achieved by writing dedicated parsing algorithms or using existing task planning software. These algorithms or software can usually analyze the requirements of the tasks, the capabilities of the robotic arms, and the constraints of the working environment, and then generate one or more feasible task execution plans; when generating the plans, it is also necessary to consider the cooperation relationship, conflicts, and interferences between the robotic arms.

[0064] Specifically, assume there is a dual-arm robot working in an automated warehouse. Its main task is to pick specific items from the shelves and place them into packing boxes. Before the task starts, sensors installed on the robotic arms (such as encoders, gyroscopes, etc.) are used to monitor in real-time parameters such as the position, speed, and joint angles of the robotic arms. At the same time, machine vision technology is utilized to monitor the positions of the shelves and packing boxes as well as the status of the items. The picking task is received from the Warehouse Management System (WMS), including information about the items to be picked (such as item numbers, quantities, etc.) and the position information of the target packing box.

[0065] Analyze the picking task to determine the sub-work tasks that each robotic arm needs to complete. For example, robotic arm 1 is assigned to pick items from the shelves and place them in a temporary storage area, while robotic arm 2 is responsible for picking up the items from the temporary storage area and placing them into the target packing box. To avoid conflicts and interference between the robotic arms, it is also necessary to plan their movement paths and cooperation strategies based on the layout of the shelves and packing boxes and the movement capabilities of the robotic arms. Through the above steps, the two robotic arms of the dual-arm robot are successfully monitored in real-time, the work tasks are collected, and the tasks are analyzed to determine the sub-work tasks that each robotic arm needs to complete, which provides an important guarantee for the efficient and cooperative execution of the dual-arm robot in subsequent tasks.

[0066] Therefore, determine the working paths of the two robotic arms according to the first sub-work task, the second sub-work task, and the two robotic arms of the dual-arm robot. Determine multiple anti-interference warning points based on the working paths of the two robotic arms and the movement ranges of the two robotic arms. Determine the anti-interference logic between the two robotic arms based on the multiple anti-interference warning points and the training of the two robotic arms. This anti-interference logic ensures that the two robotic arms do not come into contact with each other, takes into account the overall consideration of the multiple anti-interference warning points and the training of the two robotic arms, and ensures the accuracy of the anti-interference logic between the two robotic arms.

[0067] At this time, according to the first sub-work task, the second sub-work task, and the current states and capabilities of the two robotic arms of the dual-arm robot, determine the working paths that each of them needs to follow. The working path should be able to efficiently guide the robotic arm to move from the starting position to the target position while complying with the kinematic constraints of the robotic arm and the safety requirements of the working environment. At this time, in actual operation, the determination of the working path usually relies on path planning algorithms, which can consider various factors such as the joint limitations of the robotic arm, the positions of obstacles, and the layout of the working space, and generate one or more feasible paths. The path planning algorithms are based on graph search (such as A* algorithm, Dijkstra algorithm, etc.) and are also based on sampling (such as Rapidly-exploring Random Tree RRT, Probabilistic Roadmap PRM, etc.).

[0068] After determining the working paths of the two robotic arms, a series of anti-interference warning points need to be set along these paths. These warning points are the risk points where the robotic arms may collide with other objects (including another robotic arm) during movement. By setting sensors or monitoring mechanisms at these points, the relative positions and distances between the robotic arms can be monitored in real time to trigger avoidance actions when necessary. At this time, the determination of anti-interference warning points usually depends on a detailed analysis of the movement ranges and working paths of the robotic arms. Simulation software is used to simulate the movement of the robotic arms and calculate the collision risks at different positions. According to the simulation results, warning points are set in high-risk areas and corresponding sensors or monitoring devices are configured.

[0069] The anti-interference logic is a set of rules and strategies for ensuring that two robotic arms complete tasks without contacting each other. It should be able to monitor the positions and speeds of the robotic arms in real time, predict potential collision risks, and trigger avoidance actions when necessary. The anti-interference logic is implemented based on various algorithms and technologies, such as collision detection algorithms, motion planning algorithms, real-time control systems, etc. At this time, in actual operation, the implementation of anti-interference logic usually depends on advanced control systems and algorithms that can monitor the states of the robotic arms and changes in the surrounding environment in real time and make decisions according to preset rules and strategies. For example, when the distance between the two robotic arms is detected to be less than the safety threshold, the control system automatically adjusts the movement paths or speeds of the robotic arms to avoid collisions.

[0070] Specifically, for a dual-arm robot, robotic arm 1 is responsible for grasping parts, and robotic arm 2 is responsible for installing the parts at the designated position. According to the current positions and task requirements of robotic arm 1 and robotic arm 2, path planning algorithms are used to generate their respective working paths. For example, the path of robotic arm 1 is from the part storage area to the assembly work area, while the path of robotic arm 2 is from the assembly work area to the conveyor belt of the next process.

[0071] By analyzing the working paths and operating ranges of robotic arm 1 and robotic arm 2, potential collision risk points are determined; for example, near the assembly work area that both robotic arm 1 and robotic arm 2 need to pass through, multiple anti-interference warning points are set, and these warning points are configured to trigger an alarm or avoidance action when the distance between the robotic arms is less than a certain threshold; based on the settings of the anti-interference warning points and the movement capabilities of the robotic arms, a set of anti-interference logics is formulated; for example, when it is detected that the distance between robotic arm 1 and robotic arm 2 is less than the safety threshold, the control system automatically pauses the movement of one of the robotic arms, or adjusts its movement path to avoid collision; in addition, visual sensors and machine learning algorithms are also used to continuously monitor the relative positions and postures between the robotic arms, and dynamic avoidance decisions are made based on this information; through the above steps, the working paths, anti-interference warning points, and anti-interference logics of the two robotic arms of the dual-arm robot are successfully determined, which provides an important guarantee for the efficient and safe execution of the dual-arm robot in subsequent tasks.

[0072] In an embodiment of the present application, the dual-arm robot is performing an assembly task, where robotic arm 1 is responsible for grasping parts and robotic arm 2 is responsible for installing the parts at the designated position; there are some obstacles (such as workbenches, shelves, etc.) in the working space, as well as potential conflict areas (such as the assembly area that both robotic arm 1 and robotic arm 2 pass through); the warning point matching table is shown in Table 2: Table 2 Warning Point Matching Table Robot arm Warning point Relative position (distance, angle) Timestamp 1 A (50cm, 45°) T1 2 A (45cm, 135°) T1 + Δt1 1 B (80cm, 90°) T2 2 C (120cm, 0°) T2 + Δt2 Reference Figure 6 , in step S15, based on the moving direction of the dual-arm robot, the form of the dual-arm robot, and the operating ranges of the two robotic arms, the multi-dimensional operating range of the dual-arm robot is determined, and based on the multi-dimensional operating range of the dual-arm robot, the anti-interference logic, and the three-dimensional avoidance logic, multiple obstacle avoidance of the dual-arm robot with respect to each obstacle area is triggered; In the specific implementation process of the present invention, the specific steps are as follows: S151: Collect the moving direction of the dual-arm robot, and determine the first-dimensional operating range according to the moving direction of the dual-arm robot and the form of the dual-arm robot; S152: Determine the second-dimensional operating range of the dual-arm robot according to the moving direction of the dual-arm robot and the operating ranges of the two robotic arms, and determine the multi-dimensional operating range of the dual-arm robot based on the first-dimensional operating range, the second-dimensional operating range, and the operating range mapping relationship; S153: In the multi-dimensional operating range of the dual-arm robot, combine the anti-interference logic and the three-dimensional avoidance logic, and determine a multiple obstacle avoidance mechanism according to the combination of the anti-interference logic, the three-dimensional avoidance logic, and the moving state of the dual-arm robot, and trigger multiple obstacle avoidance of the dual-arm robot with respect to each obstacle area based on this multiple obstacle avoidance mechanism.

[0073] In an embodiment of the present application, the moving direction of the dual-arm robot is collected, and the first-dimensional activity range is determined according to the moving direction of the dual-arm robot and the form of the dual-arm robot, which takes into account both the moving direction of the dual-arm robot and the form of the dual-arm robot as a whole, ensuring the accuracy of the first-dimensional activity range.

[0074] At this time, sensors or control systems are used to monitor and record the moving direction of the dual-arm robot; the moving direction can be straight line, curve, rotation, etc., depending on the motion ability and task requirements of the dual-arm robot; the collected data includes the speed, acceleration, steering angle, etc. of the dual-arm robot, and these data are crucial for determining the activity range subsequently; at this time, in actual operation, sensors such as gyroscopes, accelerometers, and magnetometers are installed on the dual-arm robot to collect the moving direction data in real time, and these sensors can sense the attitude changes of the dual-arm robot and transmit the data to the control system for processing; in addition, machine vision technology is also used to monitor the motion trajectory of the dual-arm robot to indirectly obtain the moving direction information.

[0075] This step involves analyzing the physical form of the dual-arm robot, including the length of the robotic arm, joint configuration, size of the end effector, etc., to determine the basic activity range of the dual-arm robot in the stationary state; the basic activity range refers to the spatial area that the robotic arm of the dual-arm robot can reach without moving; at this time, determining the basic activity range usually requires using 3D modeling software or robotic arm simulation tools, and these tools can generate the reachable space diagrams of the dual-arm robot in different postures according to the geometric dimensions and kinematic parameters of the dual-arm robot; by analyzing these diagrams, the basic activity range of the dual-arm robot in the stationary state is obtained.

[0076] After obtaining the moving direction and the basic activity range of the dual-arm robot, this step involves combining the two to determine the first-dimensional activity range of the dual-arm robot during movement; the first-dimensional activity range refers to the spatial area that the dual-arm robot can reach in a specific moving direction, which takes into account the influence of the moving ability and physical form of the dual-arm robot on the reachable space; at this time, determining the first-dimensional activity range usually requires using dynamic simulation or path planning algorithms, and these algorithms can simulate the motion trajectories of the dual-arm robot in different moving directions and combine the information of the basic activity range to generate the reachable space diagrams of the dual-arm robot during movement; by analyzing these diagrams, the first-dimensional activity range of the dual-arm robot in a specific moving direction is obtained.

[0077] Specifically, assume there is a dual-arm robot working in an automated warehouse, and its main task is to pick up goods from the shelves and place them on the conveyor belt; the dual-arm robot moves from the shelf area to the conveyor belt area in a straight line; through the gyroscopes and accelerometers installed on the dual-arm robot, data such as the speed, acceleration, and steering angle of the dual-arm robot during movement are collected in real time; a three-dimensional modeling software is used to model the physical form of the dual-arm robot, including the lengths of the robotic arms, joint configurations, and the dimensions of the end effectors, etc.; through simulation analysis, the basic working range of the dual-arm robot in a stationary state is obtained, that is, the spatial area that the robotic arms can reach.

[0078] For the first-dimensional working range, a dynamic simulation algorithm or path planning software is used to simulate the movement trajectory of the dual-arm robot during straight-line movement; combined with the information of the basic working range, a reachable space map of the dual-arm robot during movement is generated; by analyzing these graphs, the first-dimensional working range of the dual-arm robot in the straight-line movement direction is obtained, that is, the area of goods that the dual-arm robot can reach during movement; through the above steps, the first-dimensional working range of the dual-arm robot in a specific movement direction is successfully determined, providing an important basis for subsequent task planning and obstacle avoidance strategies.

[0079] Furthermore, the second-dimensional working range of the dual-arm robot is determined according to the movement direction of the dual-arm robot and the working ranges of the two robotic arms, and the multi-dimensional working range of the dual-arm robot is determined based on the first-dimensional working range, the second-dimensional working range, and the working range mapping relationship, taking into account the overall consideration of the first-dimensional working range, the second-dimensional working range, and the working range mapping relationship, ensuring the accuracy of the multi-dimensional working range of the dual-arm robot.

[0080] At this time, the working ranges of the two robotic arms of the dual-arm robot in a specific movement direction are analyzed; different from the first dimension (usually referring to the horizontal or straight-line movement direction), the second-dimensional working range considers the reachable space of the dual-arm robot in the vertical direction, depth direction, or other non-straight-line directions, which usually involves a comprehensive evaluation of the joint movements of the robotic arms of the dual-arm robot, the operating range of the end effectors, and the existing obstacles; at this time, three-dimensional space analysis tools or robotic arm simulation software are needed to determine the second-dimensional working range, and these tools can simulate the robotic arm movements of the dual-arm robot in different postures and positions and calculate the spatial area that the robotic arms can reach in a specific direction; in addition, the structural limitations of the dual-arm robot itself (such as the chassis height, robotic arm joint angle limitations, etc.) and the existing obstacles in the environment also need to be considered.

[0081] After determining the activity ranges of the first dimension and the second dimension, this step involves integrating the information of these two dimensions to form a more comprehensive description of the activity range. This usually involves calculating the spatial intersection, union, or other logical combinations of the activity ranges of the two dimensions to obtain the overall activity ability of the dual-arm robot in multi-dimensional space. At this time, integrating the activity ranges of the first dimension and the second dimension requires using spatial geometry algorithms or 3D modeling software, which can handle complex three-dimensional spatial relationships and calculate the intersection, union, etc. of the activity ranges of the dual-arm robot in different dimensions. By integrating this information, the overall activity range of the dual-arm robot in multi-dimensional space is obtained.

[0082] The activity range mapping relationship refers to the process of mapping the activity ranges of the dual-arm robot in different dimensions to a unified three-dimensional space coordinate system. This step involves converting the activity ranges of the first dimension and the second dimension (as well as other dimensions) into representations in the unified coordinate system for spatial analysis and task planning. At this time, determining the multi-dimensional activity range usually requires establishing a three-dimensional space model and mapping the activity ranges of the dual-arm robot in different dimensions into this model, which requires technical means such as spatial coordinate transformation algorithms and 3D visualization tools. Through the mapping relationship, the activity ranges of the dual-arm robot in different dimensions are integrated into a unified three-dimensional space, thereby obtaining the overall activity range of the dual-arm robot in multi-dimensional space.

[0083] Specifically, for a dual-arm robot, its main task is to pick up goods from shelves at different heights and place them on the assembly line. Determine the activity range of the second dimension: The dual-arm robot needs to move vertically to reach shelves at different heights. By analyzing the joint angle limits of the robot's manipulator, the operating range of the end effector, and the height distribution of the shelves, determine the activity range of the second dimension of the dual-arm robot in the vertical direction. For example, the dual-arm robot can reach goods on shelves from the ground to 3 meters high.

[0084] The movement range of the dual-arm robot in the horizontal direction (the first dimension) is restricted by the factory layout, the position of the assembly line, and the presence of other dual-arm robots. At the same time, the movement range of the dual-arm robot in the vertical direction (the second dimension) is also restricted by the shelf height and the joint angles of the manipulator. By comprehensively evaluating the activity ranges of these two dimensions, the overall activity ability of the dual-arm robot in multi-dimensional space is obtained. For example, the dual-arm robot can move 5 meters in the horizontal direction and reach goods on shelves from the ground to 3 meters high in the vertical direction.

[0085] To more intuitively represent the working range of a dual-arm robot in multi-dimensional space, a three-dimensional space model is established, and the working ranges of the dual-arm robot in different dimensions are mapped into this model. In this model, the reachable spaces of the dual-arm robot in the horizontal and vertical directions, as well as the overlapping parts between these spaces, can be clearly seen. Through the mapping relationship, the overall working range of the dual-arm robot in multi-dimensional space is obtained, and based on this, task planning and obstacle avoidance strategies are formulated. Through the above steps, the working range of the dual-arm robot in multi-dimensional space is successfully determined, providing an important basis for subsequent task execution and obstacle avoidance strategies.

[0086] Therefore, in the multi-dimensional working range of the dual-arm robot, the anti-interference logic and the three-dimensional avoidance logic are combined, and a multiple obstacle avoidance mechanism is determined based on the combination of the anti-interference logic, the three-dimensional avoidance logic, and the moving state of the dual-arm robot. Based on this multiple obstacle avoidance mechanism, multiple obstacle avoidances of the dual-arm robot relative to each obstacle area are triggered, which takes into account the overall consideration of the multi-dimensional working range, the anti-interference logic, and the three-dimensional avoidance logic of the dual-arm robot, ensuring the intelligence of the multiple obstacle avoidances of the dual-arm robot relative to each obstacle area and improving the accuracy of obstacle avoidance of the dual-arm robot relative to each obstacle area.

[0087] At this time, the anti-interference logic (ensuring that the two manipulators of the dual-arm robot do not interfere with each other) and the three-dimensional avoidance logic (ensuring that the dual-arm robot does not collide with other obstacles in the environment) are combined. The anti-interference logic usually considers factors such as the movement trajectory of the manipulator, the joint angle, and the position of the end effector to ensure a safe distance between the manipulators. The three-dimensional avoidance logic considers the overall movement of the dual-arm robot in three-dimensional space, including translation, rotation, etc., to ensure a safe distance between the dual-arm robot and fixed obstacles, moving obstacles, etc. in the environment. At this time, technical means such as path planning algorithms, collision detection algorithms, and real-time monitoring systems are required to combine these two logics. The path planning algorithm is used to generate a safe movement path for the dual-arm robot, the collision detection algorithm is used to monitor the distance between the dual-arm robot and the obstacles in the environment in real time, and the real-time monitoring system is used to collect and process the movement state data of the dual-arm robot.

[0088] After combining the anti-interference logic and the three-dimensional avoidance logic, it is necessary to determine the multiple obstacle avoidance mechanism according to the real-time movement state of the dual-arm robot (such as position, speed, acceleration, etc.). The multiple obstacle avoidance mechanism includes different obstacle avoidance strategies, such as deceleration obstacle avoidance, detour obstacle avoidance, pause obstacle avoidance, etc. The selection of these strategies depends on factors such as the relative position, speed, and acceleration between the dual-arm robot and the obstacle. At this time, decision algorithms such as decision trees, fuzzy logic, and neural networks are required to determine the multiple obstacle avoidance mechanism. These algorithms can dynamically select appropriate obstacle avoidance strategies according to the real-time movement state of the dual-arm robot and the obstacle information.

[0089] After determining the multiple obstacle avoidance mechanism, it is necessary to trigger corresponding obstacle avoidance actions based on the real-time motion state of the dual-arm robot and obstacle information, which involves actions such as adjusting the speed of the dual-arm robot, changing its moving path, and pausing its motion to ensure that the dual-arm robot does not collide with any obstacles during the task execution; at this time, hardware devices such as motion controllers and actuators are required to trigger the multiple obstacle avoidance mechanism; the motion controller is responsible for receiving instructions from the decision-making algorithm and adjusting the motion state of the dual-arm robot according to these instructions; the actuator is responsible for converting the instructions of the motion controller into the actual motion of the dual-arm robot.

[0090] Specifically, assume that there is a dual-arm robot working in a busy automated warehouse, and its main task is to pick up goods from the shelves and place them on the conveyor belt; there are multiple shelves, conveyor belts, other dual-arm robots, and personnel and other obstacles in the warehouse; the dual-arm robot needs to consider both the anti-interference between the two robotic arms and the three-dimensional avoidance of other obstacles in the environment; for example, when the dual-arm robot moves from one shelf to another, it needs to ensure that the two robotic arms do not collide with each other, and the dual-arm robot itself does not collide with the shelves, conveyor belts, or other dual-arm robots. This requires using path planning algorithms to generate a safe moving path for the dual-arm robot and using collision detection algorithms to monitor the distance between the dual-arm robot and obstacles in real time.

[0091] Based on the real-time moving state of the dual-arm robot and obstacle information, determine the multiple obstacle avoidance mechanism; for example, when the dual-arm robot approaches a shelf, it needs to decelerate to avoid colliding with the shelf; when the dual-arm robot detects that another dual-arm robot is on its path, it needs to detour to avoid colliding with it; when the dual-arm robot detects an emergency situation ahead (such as a person suddenly entering the working area), it needs to immediately pause its motion to ensure safety.

[0092] Once the multiple obstacle avoidance mechanism is determined, it is necessary to trigger corresponding obstacle avoidance actions based on the real-time motion state of the dual-arm robot and obstacle information; for example, when the dual-arm robot needs to decelerate, the motion controller adjusts the speed of the dual-arm robot; when the dual-arm robot needs to detour, the path planning algorithm generates a new moving path; when the dual-arm robot needs to pause its motion, the actuator immediately stops all motions of the dual-arm robot; through the above steps, the anti-interference logic and three-dimensional avoidance logic are successfully combined, and the multiple obstacle avoidance mechanism is determined according to the real-time moving state of the dual-arm robot, thus ensuring the safety of the dual-arm robot during the task execution.

[0093] In one embodiment of the present application, an obstacle avoidance strategy matching table is constructed. The obstacle avoidance strategy matching table lists the obstacle avoidance strategies that the dual-arm robot should adopt in different situations. The columns of the matching table include anti-interference logic, three-dimensional avoidance logic, and the movement state of the dual-arm robot (such as speed, direction, etc.), and the rows correspond to different obstacle avoidance strategies (such as deceleration, detouring, pausing, etc.). The obstacle avoidance strategy matching table is shown in Table 3: Table 3 Obstacle Avoidance Strategy Matching Table No anti-interference requirement, no three-dimensional avoidance requirement, low-speed movement Normal driving With anti-interference requirement, no three-dimensional avoidance requirement, medium-speed movement Decelerate to avoid obstacles No anti-interference requirement, with three-dimensional avoidance requirement, high-speed movement Detour to avoid obstacles With anti-interference requirement, with three-dimensional avoidance requirement, any speed Pause to avoid obstacles Suppose that when the dual-arm robot is performing a task, it simultaneously detects the need for anti-interference and three-dimensional avoidance, and is moving at medium speed. According to the matching table, the dual-arm robot will adopt the "pause for obstacle avoidance" strategy to ensure safety.

[0094] Please refer to Figure 7 , Figure 7 which is a schematic structural diagram of the obstacle avoidance system of the dual-arm robot in an embodiment of the present invention. The obstacle avoidance system of the dual-arm robot includes: A movement path module 21, configured to determine the actual movement path of the dual-arm robot according to the current position and target position of the dual-arm robot; An obstacle area module 22, configured to determine the current positions of multiple obstacles based on the obstacle detection of the actual movement path of the dual-arm robot, and determine multiple obstacle areas according to the current positions of the multiple obstacles and the shapes of the multiple obstacles; A three-dimensional avoidance module 23, configured to determine three-dimensional avoidance logic during the movement of the dual-arm robot according to the movement position of the dual-arm robot, the spatial positions of the two robotic arms of the dual-arm robot, and multiple obstacle areas; An anti-interference logic module 24, configured to determine the anti-interference logic between the two robotic arms of the dual-arm robot according to the working paths and the activity ranges of the two robotic arms; A multiple obstacle avoidance module 25, configured to determine the multi-dimensional activity range of the dual-arm robot based on the movement direction of the dual-arm robot, the shape of the dual-arm robot, and the activity ranges of the two robotic arms, and trigger multiple obstacle avoidance of the dual-arm robot relative to each obstacle area based on the multi-dimensional activity range, anti-interference logic, and three-dimensional avoidance logic of the dual-arm robot.

[0095] Arbitrary combinations of the technical features of the above embodiments are made. To make the description more concise, not all combinations of the technical features in the above embodiments are described. However, as long as there are no technical contradictions in the combinations of these technical features, they should all be considered as the main scope recorded in this specification.

Claims

1. An obstacle avoidance method for a dual-arm robot, characterized in that: include: Determine the actual moving path of the dual-arm robot according to the current position and target position of the dual-arm robot; Determine the current positions of the plurality of obstacles based on obstacle detection of the actual moving path of the dual-arm robot, and determine the plurality of obstacle areas according to the current positions of the plurality of obstacles and the shapes of the plurality of obstacles; During the movement of the dual-arm robot, a three-dimensional avoidance logic is determined according to the moving position of the dual-arm robot, the spatial positions of the two mechanical arms of the dual-arm robot and multiple obstacle areas; In the two mechanical arms of the dual-arm robot, the anti-interference logic between the two mechanical arms is determined according to the working paths of the two mechanical arms and the activity ranges of the two mechanical arms; The multi-dimensional activity range of the dual-arm robot is determined based on the moving direction of the dual-arm robot, the shape of the dual-arm robot and the activity range of the two robotic arms, and the multiple obstacle avoidance of the dual-arm robot relative to each obstacle area is triggered based on the multi-dimensional activity range of the dual-arm robot, anti-interference logic and three-dimensional avoidance logic.

2. The obstacle avoidance method of the dual-arm robot according to claim 1, characterized in that: Determining the actual moving path of the dual-arm robot according to the current position and the target position of the dual-arm robot comprises: Collecting the current position and target position of the dual-arm robot, determining the moving space of the dual-arm robot according to the current position and target position of the dual-arm robot, and determining a plurality of alternative moving paths based on the moving space and the previous moving paths of the dual-arm robot; Determine three better moving paths based on preliminary screening of the detection of multiple candidate moving paths, and the matching coefficients of the three better moving paths and the dual-arm robot meet a preset matching coefficient threshold; Based on the detection of three optimal moving paths, the moving distance of each optimal moving path is determined, and the actual moving path of the dual-arm robot is determined according to the moving distance of each optimal moving path, the shape of the dual-arm robot and the power of the dual-arm robot.

3. The obstacle avoidance method of the dual-arm robot according to claim 1, characterized in that: The obstacle detection based on the actual moving path of the dual-arm robot determines the current positions of the multiple obstacles, and determines the multiple obstacle areas according to the current positions of the multiple obstacles and the shapes of the multiple obstacles, including: The actual moving path of the dual-arm robot is collected, and multiple obstacles in the actual moving path are marked according to the detection of the actual moving path of the dual-arm robot, and the current positions of the multiple obstacles are determined; the current positions of the multiple obstacles include the horizontal positions and height positions of the multiple obstacles; The dual-arm robot is equipped with a radar, and collects corresponding point cloud sets according to the detection of each obstacle by the radar, and determines the shape and type of the obstacle according to the recognition of each point cloud set; The activity influence range corresponding to the obstacle is determined based on the shape and type of the obstacle, and multiple obstacle areas are determined according to the horizontal positions and height positions of the multiple obstacles and the activity influence ranges corresponding to the obstacles.

4. The obstacle avoidance method of the dual-arm robot according to claim 1, characterized in that: During the movement of the dual-arm robot, the three-dimensional avoidance logic is determined according to the moving position of the dual-arm robot, the spatial positions of the two mechanical arms of the dual-arm robot and multiple obstacle areas, including: The movement of the dual-arm robot is monitored in real time, the moving position of the dual-arm robot is collected, and multiple obstacle areas around the dual-arm robot are determined based on the peripheral detection of the moving position of the dual-arm robot. At this time, the movable range of the dual-arm robot during the movement is determined based on the detection of the dual-arm robot and multiple obstacle areas around the dual-arm robot.

5. The obstacle avoidance method of the dual-arm robot according to claim 4, characterized in that: During the movement of the dual-arm robot, the three-dimensional avoidance logic is determined according to the moving position of the dual-arm robot, the spatial positions of the two mechanical arms of the dual-arm robot and the multiple obstacle areas, and further includes: In the dual-arm robot, a corresponding dual-arm robot coordinate system is constructed based on the dual-arm robot, and the spatial positions of the two mechanical arms of the dual-arm robot are determined according to the two mechanical arms of the dual-arm robot and the dual-arm robot coordinate system; The activity dimensions of the two robotic arms of the dual-arm robot are determined based on the traversal of the two robotic arms, and the activity ranges of the two robotic arms of the dual-arm robot are determined according to the activity dimensions of the two robotic arms and the relative distance between the two robotic arms, and the three-dimensional avoidance logic is determined according to the movable range of the dual-arm robot during the movement and the activity ranges of the two robotic arms of the dual-arm robot.

6. The obstacle avoidance method of a dual-arm robot according to claim 1, characterized in that: In the two mechanical arms of the dual-arm robot, the anti-interference logic between the two mechanical arms is determined according to the working paths of the two mechanical arms and the activity ranges of the two mechanical arms, including: The activity ranges of the two robotic arms are determined based on the detection of the two robotic arms of the dual-arm robot. At this time, the activity ranges of the two robotic arms are further optimized to avoid overlapping areas between the activity ranges of the two robotic arms.

7. The obstacle avoidance method of the dual-arm robot according to claim 6, characterized in that: In the two mechanical arms of the dual-arm robot, the anti-interference logic between the two mechanical arms is determined according to the working paths of the two mechanical arms and the activity ranges of the two mechanical arms, and further includes: Monitor the two mechanical arms of the dual-arm robot in real time, collect the work tasks of the dual-arm robot, and determine the sub-work tasks corresponding to the two mechanical arms according to the analysis of the work tasks, where the sub-work tasks corresponding to the two mechanical arms are the first sub-work task and the second sub-work task respectively; The working paths of the two robotic arms are determined according to the first sub-work task, the second sub-work task and the two robotic arms of the dual-arm robot, a plurality of anti-interference warning points are determined according to the working paths of the two robotic arms and the activity range of the two robotic arms, and the anti-interference logic between the two robotic arms is determined based on the plurality of anti-interference warning points and the training of the two robotic arms, and the anti-interference logic ensures that the two robotic arms do not contact each other.

8. The obstacle avoidance method of a dual-arm robot according to claim 1, characterized in that: The method of determining the multi-dimensional activity range of the dual-arm robot based on the moving direction of the dual-arm robot, the form of the dual-arm robot and the activity range of the two robotic arms, and triggering the dual-arm robot to avoid multiple obstacles relative to each obstacle area based on the multi-dimensional activity range of the dual-arm robot, anti-interference logic and three-dimensional avoidance logic, includes: Collect the moving direction of the dual-arm robot, and determine the first-dimensional activity range according to the moving direction and shape of the dual-arm robot; The second dimensional activity range of the dual-arm robot is determined according to the moving direction of the dual-arm robot and the activity range of the two robotic arms, and the multi-dimensional activity range of the dual-arm robot is determined based on the first dimensional activity range, the second dimensional activity range and the activity range mapping relationship.

9. The obstacle avoidance method of the dual-arm robot according to claim 8, characterized in that: The method further includes determining the multi-dimensional activity range of the dual-arm robot based on the moving direction of the dual-arm robot, the form of the dual-arm robot and the activity range of the two mechanical arms, and triggering the dual-arm robot to avoid multiple obstacles relative to each obstacle area based on the multi-dimensional activity range of the dual-arm robot, the anti-interference logic and the three-dimensional avoidance logic. In the multi-dimensional activity range of the dual-arm robot, the anti-interference logic and the three-dimensional avoidance logic are combined, and a multiple obstacle avoidance mechanism is determined based on the combination of the anti-interference logic, the three-dimensional avoidance logic and the moving state of the dual-arm robot. Based on the multiple obstacle avoidance mechanism, the dual-arm robot triggers multiple obstacle avoidance relative to each obstacle area.

10. An obstacle avoidance system for a dual-arm robot, characterized in that: The obstacle avoidance system of the dual-arm robot is applied to the obstacle avoidance method of the dual-arm robot as claimed in any one of claims 1 to 9, and the obstacle avoidance system of the dual-arm robot comprises: A moving path module is used to determine the actual moving path of the dual-arm robot according to the current position and target position of the dual-arm robot; An obstacle area module, used to determine the current positions of multiple obstacles based on obstacle detection of an actual moving path of the dual-arm robot, and determine multiple obstacle areas according to the current positions of the multiple obstacles and the shapes of the multiple obstacles; A three-dimensional avoidance module is used to determine the three-dimensional avoidance logic according to the moving position of the dual-arm robot, the spatial positions of the two mechanical arms of the dual-arm robot and multiple obstacle areas during the movement of the dual-arm robot; The anti-interference logic module is used to determine the anti-interference logic between the two mechanical arms of the dual-arm robot according to the working paths of the two mechanical arms and the activity ranges of the two mechanical arms; The multiple obstacle avoidance module is used to determine the multi-dimensional activity range of the dual-arm robot based on the moving direction of the dual-arm robot, the shape of the dual-arm robot and the activity range of the two robotic arms, and trigger the multiple obstacle avoidance of the dual-arm robot relative to each obstacle area based on the multi-dimensional activity range of the dual-arm robot, anti-interference logic and three-dimensional avoidance logic.

Images