Path planning method, device and equipment for multi-robot cooperative processing and medium

Abstract

The invention relates to the technical field of path planning, and discloses a path planning method and device for multi-robot cooperative processing, equipment and a medium. The method comprises the following steps: acquiring environment data of a path planning stage; according to the environment data of the path planning stage, updating the current digital twinborn body to obtain an updated digital twinborn body; task allocation and path planning based on a task allocation result are carried out on the robot in the updated digital twin, and an initial path of the robot is generated; and performing conflict risk analysis and path adjustment based on a conflict risk analysis result on the initial path of the robot in the updated digital twin, and generating a task path of the robot. According to the embodiment of the invention, the path of the robot is dynamically optimized based on the digital twinning technology, so that the path obtained by dynamic planning can adapt to a dynamically changing scene.

Description

Technical Field

[0001] The present application relates to the field of path planning technology, and in particular to a path planning method, device, equipment and medium for multi-robot collaborative processing. Background Art

[0002] The autonomous navigation and obstacle avoidance capabilities of robots in complex environments are often limited by factors such as sensor accuracy, computing resources, and environmental dynamics. Among related technologies, digital twin real-time simulation and virtual environment reconstruction can provide robots with new solutions.

[0003] However, while digital twin technology offers significant advantages for robot path planning, achieving this goal still faces numerous challenges. For example, real-time processing and integration of massive amounts of data from multiple sensors presents difficulties, leading to delayed system responses, insufficient accuracy in the generated 3D models, which impacts the robot's environmental perception capabilities, and poor synchronization between the physical entity and its virtual twin. Summary of the Invention

[0004] The purpose of this application is to provide a path planning method, device, equipment and medium for multi-robot collaborative processing, which dynamically optimizes the robot's path based on digital twin technology, so that the path obtained by dynamic planning can adapt to dynamically changing scenarios.

[0005] The present invention provides a method for path planning in multi-robot collaborative processing, including: Acquire environmental data during the path planning phase; the environmental data includes environmental state data, robot state data, and robot task state data; Updating the current digital twin based on the environmental data from the path planning phase to obtain an updated digital twin; the digital twin is obtained by digitally mapping the task environment and multiple robots; Performing task assignment on the robot and path planning based on the task assignment results in the updated digital twin to generate an initial path for the robot; In the updated digital twin, a conflict risk analysis is performed on the initial path of the robot and a path adjustment is performed based on the conflict risk analysis result to generate a task path for the robot.

[0006] In some embodiments, before updating the current digital twin according to the environmental data in the path planning stage, the method further includes: Obtain environmental data for the simulation modeling phase; Performing geometric modeling and physical simulation on the task environment, the robot, and task objects related to the robot task based on the environmental state data and the robot state data of the environmental data in the simulation modeling stage to obtain a corresponding simulation modeling model; The motion constraint data of the robot and the robot task state data of the environmental data in the simulation modeling stage are configured in the simulation modeling model to generate the digital twin.

[0007] In some embodiments, updating the current digital twin based on the environmental data in the path planning stage includes: Acquiring environmental deviation data; the environmental deviation data represents the deviation between the environmental data of the path planning stage and the environmental data of the simulation modeling stage; When the environmental deviation data exceeds the environmental deviation threshold data, the current digital twin is updated according to the environmental deviation data to obtain the updated digital twin.

[0008] In some embodiments, the assigning tasks to the robot in the updated digital twin and planning a path based on the task assignment results include: With the goal of minimizing a first execution cost, the robot is assigned a task based on the robot task state data to determine the robot task to be performed by the robot; the first execution cost is obtained by fitting the execution time cost of the total task and the task conflict cost; For the robot task to be performed by the robot, with the goal of minimizing the second execution cost, a reinforcement learning method is used in the updated digital twin to plan the path of the robot to generate the initial path of the robot; the second execution cost is obtained by fitting the path length cost, the execution time cost of a single task, the task conflict probability cost and the task error cost.

[0009] In some embodiments, the assigning tasks to the robot in the updated digital twin and planning a path based on the task assignment results further includes: Simulating, in the updated digital twin, the robot performing a corresponding robot task along the initial path to determine corresponding simulated task error data and a number of simulated task conflicts; When the simulated task error data exceeds the task error threshold data and / or the simulated task conflict number exceeds the task conflict threshold number, return to the step of using the reinforcement learning method to plan the path of the robot in the updated digital twin.

[0010] In some embodiments, performing a conflict risk analysis on the initial path of the robot in the updated digital twin and adjusting the path based on the conflict risk analysis results include: calculating a minimum distance between the robots when performing the robot task along the initial path; When the minimum distance is less than a preset threshold distance or there is a task conflict between the robot tasks, the initial path is adjusted in the updated digital twin according to the priority of the robot tasks to generate the task path of the robot.

[0011] In some embodiments, the multi-robot collaborative processing path planning method further includes: Acquire task error data; the task error data represents the deviation between the actual execution result and the expected execution result when the robot performs the robot task; When the task error data exceeds the task error threshold data, the current digital twin is updated according to the task error data to obtain the updated digital twin.

[0012] The present application also provides a path planning device for multi-robot collaborative processing, including: The first module is used to obtain environmental data in the path planning stage; the environmental data includes environmental state data, robot state data and robot task state data; The second module is used to update the current digital twin based on the environmental data in the path planning stage to obtain an updated digital twin; the digital twin is obtained by digitally mapping the task environment and multiple robots; A third module is configured to perform task assignment for the robot in the updated digital twin and perform path planning based on the task assignment results to generate an initial path for the robot; The fourth module is used to perform conflict risk analysis on the initial path of the robot in the updated digital twin and adjust the path based on the conflict risk analysis results to generate the task path of the robot.

[0013] An embodiment of the present application further provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the above-mentioned path planning method for multi-robot collaborative processing when executing the computer program.

[0014] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned path planning method for multi-robot collaborative processing is implemented.

[0015] The beneficial effects of the present application are as follows: by pre-modeling the task environment as a digital twin, updating the current digital twin in the path planning phase based on the environmental data in the path planning phase, then assigning tasks to the robot and performing path planning based on the task assignment results in the updated digital twin to generate the robot's initial path, performing conflict risk analysis on the robot's initial path in the updated digital twin and adjusting the path based on the conflict risk analysis results to generate the robot's task path. Since the current digital twin is updated in the path planning phase based on the environmental data in the path planning phase, the updated digital twin maintains consistency with the current task environment, so that the robot's task path planned in the updated digital twin can adapt to dynamically changing scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a diagram of the application environment of the path planning method for multi-robot collaborative processing provided in an embodiment of the present application.

[0017] Figure 2 This is a flow chart of a path planning method for multi-robot collaborative processing provided in an embodiment of the present application.

[0018] Figure 3 It is a structural diagram of a path planning device for multi-robot collaborative processing provided in an embodiment of the present application.

[0019] Figure 4 This is a schematic diagram of the hardware structure of the electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0021] It should be noted that although the device schematics illustrate functional module divisions and the flowcharts illustrate logical sequences, in certain circumstances, the steps illustrated may be performed in a sequence that differs from the module divisions in the device or the sequence in the flowcharts. Terms such as "first" and "second" in the specification, claims, and drawings are used to distinguish similar items and are not intended to describe a specific sequence or precedence.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used herein are for the purpose of describing the embodiments of this application only and are not intended to limit this application.

[0023] The information, data, and signals involved in the embodiments of this application are all authorized by the relevant objects or fully authorized by all parties, and the collection, use, and processing of relevant data comply with the relevant laws, regulations, and standards of the relevant countries and regions.

[0024] Figure 1 This is an application environment diagram of the path planning method for multi-robot collaborative processing provided by the embodiment of the present application. Figure 1 The path planning method for multi-robot collaborative processing is applied to a robot path planning system. The robot path planning system includes a terminal 110 and a server 120. The terminal 110 and the server 120 are connected via a network. The terminal 110 can be a desktop terminal, a mobile terminal, or a vehicle-mounted terminal. The mobile terminal can be at least one of a mobile phone, a tablet computer, a laptop computer, etc. The server 120 can be implemented as an independent server or a server cluster consisting of multiple servers. The terminal 110 is used to send environmental data from the path planning phase to the server 120. The server 120 is used to obtain the environmental data from the path planning phase, update the current digital twin based on the environmental data from the path planning phase, obtain an updated digital twin, assign tasks to the robots and plan paths based on the task assignment results in the updated digital twin, generate the robot's initial path, perform conflict risk analysis on the robot's initial path in the updated digital twin, and adjust the path based on the conflict risk analysis results to generate the robot's task path. The environmental data includes environmental state data, robot state data, and robot task state data. The digital twin is obtained by digitally mapping the task environment and multiple robots.

[0025] It should be understood that Figure 1 The application scenarios shown are only examples. In actual applications, the path planning method for multi-robot collaborative processing provided by the embodiments of the present application can also be applied to other scenarios. For example, the above-mentioned path planning method for multi-robot collaborative processing can be directly applied to terminal 110, and terminal 110 is used to obtain environmental data in the path planning stage, update the current digital twin based on the environmental data in the path planning stage, obtain an updated digital twin, perform task assignments on the robots and perform path planning based on the task assignment results in the updated digital twin, generate the robot's initial path, perform conflict risk analysis on the robot's initial path in the updated digital twin, and adjust the path based on the conflict risk analysis results to generate the robot's task path. Wherein, the environmental data includes environmental state data, robot state data, and robot task state data, and the digital twin is obtained by digitally mapping the task environment and multiple robots.

[0026] In an embodiment of the present application, the execution entity of the path planning method for multi-robot collaborative processing is a terminal or a server.

[0027] Figure 2 This is a flow chart of the path planning method for multi-robot collaborative processing provided by the embodiment of the present application. Figure 2 In some embodiments, the method includes but is not limited to steps S201 to S204.

[0028] Step S201: Acquire environmental data for the path planning stage.

[0029] Environmental data includes environmental status data, robot status data and robot task status data.

[0030] Environmental state data refers to dynamic information about the machining environment and obstacle locations collected in real time via a sensor network. This data can be achieved by using lidar to scan the workshop layout, vision sensors to capture workpiece position and posture changes, and infrared temperature sensors to monitor equipment operating status. It reflects parameters such as the physical distribution of obstacles, the location of the workpiece, and the ambient temperature. Robot state data refers to the real-time motion parameters of the robot itself. This data can be obtained by encoders capturing joint angles, inertial measurement units collecting acceleration information, and torque sensors measuring end-load data. It characterizes the robot's motion speed, posture, and load status. Robot task state data refers to task execution progress, resource utilization, and position coordinate information related to the robot task (for example, the coordinates of the machining point in a machining task). This data can be described using task queue length, the number of completed processes, and the number of currently occupied tools. Furthermore, by identifying and determining the corresponding position coordinates of robot task-related locations from lidar radar data, this data can be used to track the task allocation status and resource usage conflicts of each robot.

[0031] The execution entity acquires environmental data during the path planning phase from multimodal sensors monitoring the task environment (e.g., a workshop). This environmental data is transmitted via a high-performance industrial network, DDS, ensuring high-frequency environmental data collection to meet the real-time requirements of multi-robot collaborative machining.

[0032] Step S202: Update the current digital twin based on the environmental data in the path planning stage to obtain an updated digital twin.

[0033] A digital twin is a digital mapping of the task environment and multiple robots. Before a robot mission begins, multimodal sensors deployed within the task environment collect environmental data for the simulation modeling phase. This collected environmental data is then used to construct a virtual simulation model, encompassing the task scenario, workpieces, obstacles, and multiple robots—the digital twin.

[0034] Digital twin updating refers to the synchronous calibration of real-time collected environmental data with the current digital twin. Specifically, this can be achieved by aligning the coordinate systems of the physical space and the virtual space through coordinate transformation algorithms, correcting sensor data delays using interpolation compensation methods, and eliminating multi-source information conflicts using data fusion technology to ensure consistency between the virtual environment and the physical environment.

[0035] When multiple robots start to perform a task, the executing entity will synchronize the environmental data obtained in the path planning stage with the current digital twin, align the coordinate systems of the physical space and the virtual space through the coordinate transformation algorithm, use the interpolation compensation method to correct the sensor data delay, and use data fusion technology to eliminate multi-source information conflicts, so that the updated digital twin is consistent with the current task environment.

[0036] Step S203: Assign tasks to the robot in the updated digital twin and plan the path based on the task assignment results to generate an initial path for the robot.

[0037] Task allocation refers to the dynamic adjustment of robot task sequences based on real-time task requirements. Specifically, this can be achieved by using the Hungarian algorithm to solve the optimal task matching matrix, allocating urgent tasks based on an auction mechanism, and using game theory models to coordinate the competitive relationship among multiple robots, thereby balancing task execution efficiency and resource utilization.

[0038] After obtaining the updated digital twin, the execution entity assigns tasks to the robots in the updated digital twin based on the robot task status data from the environmental data in the path planning phase. This may involve assigning robot tasks to each robot using an improved contract network protocol based on the number and priority of robot tasks to obtain the corresponding task assignment results. After assigning the corresponding robot tasks to each robot, the robot path is planned based on the environmental status data and robot status data from the path planning phase. This may involve determining the robot's position information, the robot task's position information, and the position information of obstacles in the task environment based on the environmental status data and robot status data from the path planning phase. This position information is then combined with the path generation algorithm to plan the robot's path to generate the robot's initial path.

[0039] Step S204: performing a conflict risk analysis on the initial path of the robot and adjusting the path based on the conflict risk analysis results in the updated digital twin to generate a task path for the robot.

[0040] Conflict risk analysis refers to the prediction of interactive conflicts that may occur during path execution. It can be achieved by calculating the spatial overlapping areas of the robot's motion trajectory, evaluating the overlap of the time windows of each path segment, and establishing a potential field model to detect areas where repulsion exceeds the limit, thereby identifying potential collision risks and task timing conflicts.

[0041] After generating the robots' initial paths, the execution agent simulates the robots' movements along these paths within the updated digital twin. Using a collision detection algorithm, the execution agent identifies intersections between the robots' initial paths and predicts potential robot collisions. When a collision risk is detected, such as two robots entering a narrow passageway at the same time, the execution agent replans the waiting point for one of the robots based on the urgency of the task, inserting avoidance pause instructions to generate an optimized, conflict-free path—the robot's task path.

[0042] In some embodiments, before updating the current digital twin based on the environmental data of the path planning stage, it also includes: obtaining the environmental data of the simulation modeling stage; performing geometric modeling and physical simulation on the task environment, robot and task items related to the robot task based on the environmental status data and robot status data of the environmental data of the simulation modeling stage to obtain a corresponding simulation modeling model; configuring the robot's motion constraint data and the robot task status data of the environmental data of the simulation modeling stage in the simulation modeling model to generate a digital twin.

[0043] Geometric modeling is the process of converting the task environment, robot, and task items related to the robot's task into a virtual three-dimensional model. This can be achieved by topologically reconstructing the scanned data using computer-aided design software. Its function is to provide a visualization and interactive foundation for physical simulation. Physical simulation refers to the simulation of the mechanical properties and motion laws of objects in the real environment. This can be achieved by assigning physical properties such as mass and friction coefficient to the model using finite element analysis tools. Its function is to provide dynamic constraints close to the real scene for subsequent path planning. Motion constraint data refers to the set of parameters that limit the robot's range of motion. Specifically, it can include data such as joint angle limits, maximum speed, and acceleration thresholds. Its function is to provide feasibility boundary conditions for path planning.

[0044] Specifically, in the simulation modeling stage, first, environmental data is collected through multimodal sensors deployed in the task environment to obtain environmental data for the simulation modeling stage. Subsequently, based on the environmental data from the simulation modeling stage, geometric modeling is performed on the equipment, robots, task items related to the robot task, and obstacles in the task environment. For example, a virtual workshop model is built in the Unity engine, and the robot URDF file is imported to describe its mechanical structure. Next, the model is physically simulated, such as adding gravity, collision detection rules, and material properties to the model in the Gazebo simulation platform to simulate mechanical interactions in the real environment. After the modeling is completed, the robot's motion constraint data is written into the model configuration file, such as setting the maximum extension radius of the robotic arm to avoid collisions with obstacles. At the same time, the task status data is embedded in the model logic layer, such as marking the processing priority of different workpieces. The resulting digital twin can provide a high-precision virtual test environment for subsequent path planning.

[0045] In one specific embodiment, geometric modeling of the task environment, the robot, and the task objects associated with the robot's task is performed by generating a simulation model of the task environment in virtual simulation software, then importing the CAD models of the robot and the task objects associated with the robot's task into the simulation model of the task environment, representing the positions of the robot and the task objects associated with the robot's task in the task environment using a triangulated network, and then generating a dynamic mesh, such as through Euclidean clustering, to represent obstacles in the task environment. Physical simulation of the task environment, the robot, and the task objects associated with the robot's task is performed by first defining a dynamic model of the robot based on the Lagrange equations, then modeling the physical properties of the task objects associated with the robot's task (e.g., material density, thermal conductivity, and surface roughness), and then associating these physical properties with the simulation model of the task objects associated with the robot's task to establish a corresponding processing resistance model. The robot task status data of the robot's motion constraint data and the environmental data of the simulation modeling stage are configured in the simulation modeling model. Specifically, the robot task set is defined in the simulation modeling model. The robot task set contains multiple robot tasks and corresponding task timing information, and defines the geometric constraints of the robot to perform the robot tasks, including task space boundaries and collision boundaries, etc., and finally generates the corresponding digital twin.

[0046] In some embodiments, updating the current digital twin based on the environmental data in the path planning stage includes: obtaining environmental deviation data; when the environmental deviation data exceeds the environmental deviation threshold data, updating the current digital twin based on the environmental deviation data to obtain an updated digital twin.

[0047] Environmental deviation data characterizes the deviation between the environmental data in the path planning phase and the environmental data in the simulation modeling phase. It is understandable that environmental deviation data refers to the difference between the real-time environmental parameters collected in the path planning phase and the baseline model parameters established in the simulation modeling phase. Specifically, it can be achieved by using a laser rangefinder to obtain spatial coordinate deviations and a pressure sensor to obtain load change data, and is used to quantify the degree of deviation between the actual working conditions and the virtual model. Environmental deviation threshold data refers to the pre-set allowable environmental parameter fluctuation range, which can be specifically achieved by using the maximum position offset allowed in industrial scenarios or the equipment operating parameter safety interval, and is used to determine whether the model update mechanism needs to be initiated.

[0048] During the robot's execution of its tasks, the executing entity determines the environmental deviation data by comparing the environmental data from the path planning phase with the environmental data from the simulation modeling phase. For example, in the path planning phase, the workpiece position coordinates are detected by a visual sensor, and then compared with the workpiece position coordinates detected by the visual sensor established in the simulation modeling phase to calculate the offset of the workpiece position coordinates. When it is detected that the environmental deviation data exceeds the environmental deviation threshold data, the digital twin update program will be triggered to update the current digital twin based on the environmental deviation data to obtain an updated digital twin. For example, the environmental deviation threshold data can be set to a 5% position deviation or a 10% sensor data deviation. At this time, the coordinates of the workpiece model in the digital twin are corrected based on the workpiece position measurement data from the path planning phase, and the robot motion trajectory planning parameters are adjusted at the same time to ensure that the updated digital twin remains dynamically consistent with the actual physical environment.

[0049] In a specific embodiment, the expression for updating the current digital twin is: , , in, For the updated digital twin, For the current digital twin, To update the weight parameters, is the environmental deviation data, is the environmental data for the path planning stage, Environmental data for the simulation modeling phase.

[0050] In some embodiments, task assignment and path planning based on the task assignment results are performed on the robot in the updated digital twin, including: assigning tasks to the robot based on the robot task status data with the goal of minimizing the first execution cost to determine the robot task to be performed by the robot; for the robot task to be performed by the robot, with the goal of minimizing the second execution cost, path planning is performed on the robot using a reinforcement learning method in the updated digital twin to generate an initial path for the robot.

[0051] The first execution cost is calculated by fitting the total task execution time cost and the task conflict cost. It can be understood that the first execution cost is a comprehensive cost indicator used to evaluate multi-robot task allocation. Specifically, it can be achieved by linearly combining the total task execution time and the number of task conflicts using a weighted summation method. For example, the time cost coefficient is set to 0.6 and the conflict cost coefficient is set to 0.4. Its purpose is to quantify the comprehensive cost of task allocation and avoid the increased risk of robot task overlap caused by simply optimizing time.

[0052] The second execution cost is calculated by fitting the path length cost, the execution time cost of a single task, the task conflict probability cost, and the task error cost. It can be understood that the second execution cost is a comprehensive cost indicator used to evaluate the path planning of a single robot. Specifically, it can be achieved by dynamically weighting path length, task execution time, conflict probability, and operation error using a multi-objective optimization algorithm. Its role is to simultaneously constrain path efficiency, safety, and task accuracy, preventing local optimization from causing overall system performance degradation.

[0053] Reinforcement learning method refers to a path planning strategy optimization method based on environmental interaction. Specifically, it can be implemented by combining a deep Q-network algorithm with the simulation feedback data of a digital twin. Its function is to adapt to environmental changes and reduce the complexity of manual parameter adjustment by dynamically adjusting the path decision strategy.

[0054] During the task allocation phase, the robot task status data is input into the allocation model, which traverses all possible task combinations and calculates the corresponding first execution costs to select the task allocation plan with the lowest total execution cost. For example, when three robots need to complete five robot tasks, the execution entity will exclude allocation combinations with overlapping task times and give priority to the plan with the shortest total time and zero conflicts. During the path planning phase, for each robot task assigned, the reinforcement learning agent simulates the movement process in the digital twin and iteratively updates the path decision strategy so that the second execution cost gradually converges to the minimum value. For example, when the robot needs to bypass a dynamic obstacle, the agent will try different steering angles and select the plan with a conflict probability of less than 5% and the smallest path length increment.

[0055] In a specific embodiment, each robot task in a robot task set is assigned to a corresponding robot, and then a first execution cost corresponding to each task assignment scheme is calculated. Using a reinforcement learning method to plan a robot path specifically defines a corresponding state space, action space, and reward function, wherein the state space includes model data of the digital twin, robot state data, robot task state data, and robot task assignment data; the action space includes robot joint angle data; and the reward function is used to calculate a second execution cost. After constructing the state space and action space, the robot path is iteratively generated using a reinforcement learning method, and the corresponding second execution cost is calculated using the reward function, so that the second execution cost gradually converges to a minimum value to generate the robot's initial path.

[0056] The formula for calculating the first execution cost with the goal of minimizing the first execution cost is: , in, is the total task execution time cost, is the task conflict cost between the i-th robot and the j-th robot, is the cost weight parameter.

[0057] The expression of the reward function is: , in, is the second execution cost, is the path length cost, is the execution time cost of a single task, is the task conflict probability cost, is the task error cost, 、 、 and are weight coefficients respectively.

[0058] In some embodiments, the robot is assigned tasks and path planning based on the task assignment results in the updated digital twin, and further includes: simulating the robot to perform corresponding robot tasks along the initial path in the updated digital twin to determine the corresponding simulated task error data and the number of simulated task conflicts; when the simulated task error data exceeds the task error threshold data and / or the number of simulated task conflicts exceeds the task conflict threshold number, returning to the step of using the reinforcement learning method to plan the path of the robot in the updated digital twin.

[0059] The simulated task error data refers to the deviation between the actual trajectory and the planned trajectory when the robot performs the robot task. Specifically, it can be achieved by using a laser ranging sensor or a visual positioning system to collect the difference between the actual coordinates and the planned coordinates. This data is used to quantify the path execution accuracy. The task error threshold data refers to the maximum allowable trajectory deviation range, which can be set to, for example, ±2 mm or ±5 mm according to the processing accuracy requirements. The number of simulated task conflicts refers to the number of times that different robots have trajectory overlap or resource competition during the simulation process in the digital twin. Specifically, it can be achieved by counting the number of conflict events within the time window through the collision detection algorithm. The task conflict threshold number refers to the maximum number of conflicts allowed, which can be set according to the complexity of the task scenario, for example, no more than 3 conflicts in a single task.

[0060] After the initial path of the robot is initially generated, the entire process of the initial path is simulated through the digital twin, and the robot's motion trajectory data in the virtual environment is collected in real time. When it is detected that the trajectory deviation exceeds the task error threshold data or frequent collisions occur, the path replanning mechanism is automatically triggered. This process forms a closed-loop feedback system, in which the simulated error data reflects the feasibility of path planning, and the number of conflicts represents the effectiveness of multi-machine collaboration. For example, in a robotic arm welding task, if the simulation shows that the end effector deviates from the center of the weld by more than 1.5 mm, the path parameter adjustment is triggered. When the paths of two robots intersect more than three times in the material handling area, the transportation priority is reallocated and an avoidance path is planned.

[0061] In some embodiments, a conflict risk analysis is performed on the initial path of the robot in the updated digital twin, and the path is adjusted based on the conflict risk analysis results, including: calculating the minimum distance between robots when performing robot tasks along the initial path; when the minimum distance is less than a preset threshold distance or there is a task conflict between robot tasks, the initial path is adjusted in the updated digital twin according to the priority of the robot tasks to generate the robot's task path.

[0062] The preset threshold distance is the pre-set lower limit of the robot's safe operating distance. It can be set based on the robot's size, speed, and braking performance, for example, twice the robot's radius. This feature is used to determine whether to initiate a path adjustment. Priority refers to the execution order of different robot tasks. It can be sorted based on task urgency, process dependencies, or task value. For example, emergency repair tasks have higher priority than routine transport tasks. This feature is used to determine the order in which tasks are prioritized during path adjustments.

[0063] In the updated digital twin, each robot's position and trajectory are acquired. The robot's operating state on the initial path is simulated using a collision detection algorithm, and the minimum distance between any two robots is continuously calculated. When the minimum distance between robots falls below a preset threshold, a collision risk is indicated. The task execution sequence is also checked. If multiple robot tasks compete for resources or conflict in time and space, a task conflict is determined. In these situations, the execution entity replans the path based on preset priority rules, for example, forcing low-priority robots to wait or take a detour, while retaining the optimal path for high-priority robots. The adjusted path is then re-imported into the updated digital twin for simulation verification until the conflict-free condition is met and output as the final task path.

[0064] In some embodiments, the multi-robot collaborative machining path planning method further includes: obtaining task error data; and when the task error data exceeds a task error threshold, updating the current digital twin based on the task error data to obtain an updated digital twin. The task error data represents the deviation between an actual execution result and an expected execution result when the robot performs the robot task.

[0065] When the robot performs a robotic task, a high-precision encoder installed at the end of the robotic arm records the actual motion trajectory in real time and compares it with the expected trajectory in the digital twin. If the cumulative error of the actual trajectory from the expected trajectory exceeds the task error threshold data, for example, if the end of the robotic arm deviates from the expected coordinates by more than the threshold range within three consecutive motion cycles, the digital twin update process is triggered. At this time, the execution entity reversely optimizes the physical simulation parameters of the digital twin based on the error data, such as correcting the positioning deviation of the workpiece in the environmental model or adjusting the transmission error compensation coefficient of the robot joint. The updated digital twin regenerates path planning instructions and corrects the robot's motion trajectory through control instructions, thereby eliminating systematic errors caused by equipment wear or environmental interference during the actual processing process.

[0066] In a specific embodiment, the expression for updating the current digital twin is: , in, To update the weight parameters, is the task error threshold data.

[0067] See also Figure 3 The present application also provides a path planning device for multi-robot collaborative processing, which can implement the above-mentioned path planning method for multi-robot collaborative processing. The device includes: The first module 301 is used to obtain environmental data in the path planning stage; the environmental data includes environmental state data, robot state data and robot task state data; The second module 302 is used to update the current digital twin based on the environmental data in the path planning stage to obtain an updated digital twin; the digital twin is obtained by digitally mapping the task environment and multiple robots; The third module 303 is used to assign tasks to the robot in the updated digital twin and plan a path based on the task assignment results to generate an initial path for the robot; The fourth module 304 is used to perform a conflict risk analysis on the initial path of the robot in the updated digital twin and adjust the path based on the conflict risk analysis results to generate a task path for the robot.

[0068] The specific implementation of the path planning device for multi-robot collaborative processing is basically the same as the specific embodiment of the path planning method for multi-robot collaborative processing described above, and will not be repeated here.

[0069] Figure 4 It is a block diagram of an electronic device according to an exemplary embodiment.

[0070] Refer to the following Figure 4 4 to describe the electronic device 400 according to this embodiment of the present disclosure. Figure 4 The electronic device 400 shown is merely an example and should not limit the functionality and scope of use of the embodiments of the present disclosure.

[0071] like Figure 4 As shown, electronic device 400 is implemented as a general-purpose computing device. Components of electronic device 400 may include, but are not limited to, at least one processing unit 410, at least one storage unit 420, a bus 430 connecting various system components (including storage unit 420 and processing unit 410), a display unit 440, and the like.

[0072] The storage unit stores program code, which can be executed by the processing unit 410, so that the processing unit 410 executes the steps described in the path planning method part of the multi-robot collaborative processing in the above description of this specification according to various exemplary embodiments of the present disclosure.

[0073] The storage unit 420 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 4201 and / or a cache memory unit 4202 , and may further include a read-only memory unit (ROM) 4203 .

[0074] The storage unit 420 may also include a program / utility 4204 having a set (at least one) of program modules 4205, such program modules 4205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0075] Bus 430 may represent one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0076] The electronic device 400 may also communicate with one or more external devices 400′ (e.g., a keyboard, a pointing device, a Bluetooth device, etc.), one or more devices that enable a user to interact with the electronic device 400, and / or any device that enables the electronic device 400 to communicate with one or more other computing devices (e.g., a router, a modem, etc.). Such communication may occur via an input / output (I / O) interface 450. Furthermore, the electronic device 400 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network such as the Internet) via a network adapter 460. The network adapter 460 may communicate with other modules of the electronic device 400 via the bus 430. It should be understood that, although not shown in the figures, other hardware and / or software modules may be used in conjunction with the electronic device 400, including but not limited to microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0077] An embodiment of the present application further provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned path planning method for multi-robot collaborative processing.

[0078] The path planning method, apparatus, device, and medium for multi-robot collaborative processing provided in the embodiments of the present application pre-model the task environment as a digital twin. During the path planning phase, the current digital twin is updated based on the environmental data from the path planning phase. Tasks are then assigned to the robots in the updated digital twin, and path planning is performed based on the task assignment results to generate the robots' initial paths. Conflict risk analysis is then performed on the robots' initial paths in the updated digital twin, and path adjustments are made based on the conflict risk analysis results to generate the robots' task paths. Because the current digital twin is updated based on the environmental data from the path planning phase during the path planning phase, the updated digital twin maintains consistency with the current task environment, allowing the robot's task path planned in the updated digital twin to adapt to dynamically changing scenarios.

[0079] Through the description of the above embodiments, it is easy for those skilled in the art to understand that the example embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, or a network device, etc.) to execute the above-mentioned method according to the embodiments of the present disclosure.

[0080] The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or component, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0081] Computer-readable storage media may include a data signal propagated in baseband or as part of a carrier wave, which carries readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The readable storage medium may also be any readable medium other than a readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination thereof.

[0082] Those skilled in the art will appreciate that the modules described above can be distributed in the device according to the description of the embodiment, or can be modified accordingly to be used in one or more devices that are different from the embodiment. The modules of the above embodiment can be combined into one module or further divided into multiple submodules.

[0083] While the exemplary embodiments of the present disclosure have been specifically illustrated and described above, it should be understood that the present disclosure is not limited to the detailed structures, configurations, or implementations described herein; rather, the present disclosure is intended to encompass various modifications and equivalent configurations within the spirit and scope of the appended claims.

Claims

1. A path planning method for multi-robot collaborative processing, characterized in that: include: Acquire environmental data during the path planning phase; the environmental data includes environmental state data, robot state data, and robot task state data; Updating the current digital twin based on the environmental data from the path planning phase to obtain an updated digital twin; the digital twin is obtained by digitally mapping the task environment and multiple robots; Performing task assignment on the robot and path planning based on the task assignment results in the updated digital twin to generate an initial path for the robot; In the updated digital twin, a conflict risk analysis is performed on the initial path of the robot and a path adjustment is performed based on the conflict risk analysis result to generate a task path for the robot.

2. The path planning method for multi-robot collaborative processing according to claim 1, characterized in that: Before updating the current digital twin according to the environmental data in the path planning stage, the method further includes: Obtain environmental data for the simulation modeling phase; Performing geometric modeling and physical simulation on the task environment, the robot, and task objects related to the robot task based on the environmental state data and the robot state data of the environmental data in the simulation modeling stage to obtain a corresponding simulation modeling model; The motion constraint data of the robot and the robot task state data of the environmental data in the simulation modeling stage are configured in the simulation modeling model to generate the digital twin.

3. The path planning method for multi-robot collaborative processing according to claim 1, characterized in that: Updating the current digital twin based on the environmental data in the path planning stage includes: Acquiring environmental deviation data; the environmental deviation data represents the deviation between the environmental data of the path planning stage and the environmental data of the simulation modeling stage; When the environmental deviation data exceeds the environmental deviation threshold data, the current digital twin is updated according to the environmental deviation data to obtain the updated digital twin.

4. The path planning method for multi-robot collaborative processing according to claim 1, characterized in that: The performing of task assignment on the robot in the updated digital twin and path planning based on the task assignment results includes: With the goal of minimizing a first execution cost, the robot is assigned a task based on the robot task state data to determine the robot task to be performed by the robot; the first execution cost is obtained by fitting the execution time cost of the total task and the task conflict cost; For the robot task to be performed by the robot, with the goal of minimizing the second execution cost, a reinforcement learning method is used in the updated digital twin to plan the path of the robot to generate the initial path of the robot; the second execution cost is obtained by fitting the path length cost, the execution time cost of a single task, the task conflict probability cost and the task error cost.

5. The path planning method for multi-robot collaborative processing according to claim 4, characterized in that: The step of assigning tasks to the robot in the updated digital twin and planning a path based on the task assignment results further includes: Simulating, in the updated digital twin, the robot performing a corresponding robot task along the initial path to determine corresponding simulated task error data and a number of simulated task conflicts; When the simulated task error data exceeds the task error threshold data and / or the simulated task conflict number exceeds the task conflict threshold number, return to the step of using the reinforcement learning method to plan the path of the robot in the updated digital twin.

6. The path planning method for multi-robot collaborative processing according to claim 1, characterized in that: The performing of a conflict risk analysis on the initial path of the robot in the updated digital twin and adjusting the path based on the conflict risk analysis result includes: calculating a minimum distance between the robots when performing the robot task along the initial path; When the minimum distance is less than a preset threshold distance or there is a task conflict between the robot tasks, the initial path is adjusted in the updated digital twin according to the priority of the robot tasks to generate the task path of the robot.

7. The path planning method for multi-robot collaborative processing according to claim 1, characterized in that: Also includes: Obtain task error data; The task error data represents the deviation between the actual execution result and the expected execution result when the robot performs the robot task; When the task error data exceeds the task error threshold data, the current digital twin is updated according to the task error data to obtain the updated digital twin.

8. A path planning device for multi-robot collaborative processing, characterized in that: include: The first module is used to obtain environmental data in the path planning stage; the environmental data includes environmental state data, robot state data and robot task state data; The second module is used to update the current digital twin based on the environmental data in the path planning stage to obtain an updated digital twin; the digital twin is obtained by digitally mapping the task environment and multiple robots; A third module is configured to perform task assignment for the robot in the updated digital twin and perform path planning based on the task assignment results to generate an initial path for the robot; The fourth module is used to perform conflict risk analysis on the initial path of the robot in the updated digital twin and adjust the path based on the conflict risk analysis results to generate the task path of the robot.

9. An electronic device, characterized in that: The electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the path planning method for multi-robot collaborative processing according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the path planning method for multi-robot collaborative processing according to any one of claims 1 to 7 is implemented.

Images