Task cooperation management method for multiple heavy-load industrial robots
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-05
- Publication Date
- 2026-03-31
Smart Images
Figure CN121764005A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of robot task collaboration technology, and specifically to a task collaboration management method for multi-load industrial robots. Background Technology
[0002] With the continuous improvement of industrial automation, heavy-duty industrial robots are increasingly widely used in complex industrial scenarios such as logistics, assembly, and welding. In multi-robot collaborative operation scenarios, how to achieve efficient, coordinated, and safe control of multiple heavy-duty robots has become one of the key challenges in the current industrial automation field.
[0003] Currently, collaborative operations of multi-load industrial robots mostly rely on centralized scheduling or simple programmed control, lacking a deep understanding of task semantics and dynamic coordination capabilities. This makes it difficult to achieve efficient decomposition and collaborative modeling of complex tasks, resulting in low collaborative efficiency and insufficient safety. Summary of the Invention
[0004] The purpose of this invention is to provide a task collaboration management method for multi-load industrial robots to address the shortcomings in the prior art.
[0005] To achieve the above objectives, the present invention provides the following technical solution: a task collaboration management method for multi-load industrial robots, comprising the following steps: Step S1: Establish a robot collaborative cluster for several heavy-duty industrial robots, create a cluster twin model corresponding to the robot collaborative cluster using digital twin technology, and set several model collaboration points based on the distribution location of the heavy-duty industrial robots in the cluster twin model. Step S2: Set up the task and send it to the control terminal that has a communication connection with several heavy-duty industrial robots. In the control terminal, the task is processed into several sub-tasks. The sub-tasks are semantically understood and associated with each other to create a node unit, and the corresponding action collaboration unit of each sub-task is obtained. Step S3: Establish a collaborative task relationship graph between different action collaboration units. Based on the collaborative task relationship graph and the cluster twin model, perform distributed simulation control of the robot collaborative cluster in the operation simulation environment.
[0006] In a preferred embodiment, the process of establishing a robot collaborative cluster for several heavy-duty industrial robots and creating a cluster twin model corresponding to the robot collaborative cluster using digital twin technology includes: Each heavy-duty industrial robot is assigned a file interface to obtain relevant data and generate a digital file. The digital file is used to record the status and operation information of the heavy-duty industrial robot. A communication port is set up at the location of each heavy-duty industrial robot. A communication interaction layer is created based on the communication port. All communication interaction layers are connected to build a communication interaction network. The heavy-duty industrial robots exchange data based on the communication interaction network to establish a robot collaboration cluster corresponding to several heavy-duty industrial robots. Digital twin technology is used to model several heavy-duty industrial robots and their operating areas, generating corresponding operating twin spaces and robot twins. Based on the actual position coordinates of the heavy-duty industrial robots in the operating area, the corresponding robot twins are mapped to the corresponding spatial coordinates in the operating twin space to create a cluster twin model.
[0007] In a preferred embodiment, the process of setting the cluster twin model as several model collaboration points based on the distribution locations of heavy-duty industrial robots includes: Obtain the operation route of each heavy-duty industrial robot, record the intersection of different operation routes as route interconnection points, and establish cooperative relationships between different robot twins in the twin space of the same route interconnection point. Determine whether the collaboration relationship has been successfully established. If so, set the route interchange point where the collaboration relationship has been established as the model collaboration point; otherwise, re-establish the collaboration relationship.
[0008] In a preferred embodiment, the process of establishing collaborative relationships between different robot twins is as follows: Set a safe operating radius for each robot twin, obtain their respective safe operating areas based on the safe operating radius, and take the intersection area corresponding to the safe operating areas of the two robot twins as the cooperation area; Within the collaborative area, the pre-set task collaboration procedures of different robot twins are obtained. Based on the pre-set task form, all task collaboration procedures are arranged to generate a work procedure to be executed by different robot twins within the collaborative area. If the work procedure matches all the procedures of a complete work task, the collaborative relationship is successfully established; otherwise, no collaborative relationship is formed.
[0009] In a preferred embodiment, setting a task and sending it to a control terminal connected to several heavy-duty industrial robots, and then processing the task into several sub-tasks within the control terminal, includes: The system includes a control terminal with several control containers. Each control container is used to establish a transmission path for controlling the heavy-duty industrial robot and performs bidirectional security verification on the transmission path. Mark the transmission path that has passed two-way security authentication as a two-way interactive path; When in a two-way interactive path, the control terminal establishes a communication connection with and controls the heavy-duty industrial robot; when not in a two-way interactive path, the heavy-duty industrial robot is not controlled. A job task is set up, which consists of several process steps. These process steps are interdependent in execution order and have independent execution relationships. The control terminal receives the job task and processes it into branch subtasks based on the execution relationships of the process steps.
[0010] In a preferred embodiment, the process of semantically understanding and associating each branch subtask with a node unit to obtain the action collaboration unit corresponding to each branch subtask includes: A bag-of-words model is constructed to perform semantic understanding of the task content of the branch subtasks, and several task keywords of the branch subtasks are selected. Each task keyword corresponds to an action performed by the heavy-duty industrial robot. Each unique task keyword is used to construct an action matching node. All action matching nodes of the same branch subtask are aggregated, node units of the branch subtask are created, and each branch subtask is associated with the corresponding node unit. Obtain the matching relationship between each actual action performed by the heavy-duty industrial robot and the corresponding instruction keywords, set up a database to store all matching relationships, and mark the database as an action library; Import the node units of each branch subtask into the action library, filter out all the execution actions corresponding to the instruction keywords that match the task keywords from the action library, and integrate all the execution actions into the action collaboration units of the corresponding branch subtask.
[0011] In a preferred embodiment, a collaborative task relationship graph is established among different collaborative action units. Based on the collaborative task relationship graph and the cluster twin model, the process of performing distributed simulation control of the robot collaborative cluster in the work simulation environment includes: Select an action collaboration unit as the initial public graph filler object, execute all actions of the selected action collaboration unit to control the heavy-duty industrial robot, and each action is a graph filler element of the graph filler object; When the initial public graph has completed the filling of all graph filling elements, it is converted into a public task graph, which establishes the respective collaborative task graphs for other action collaboration units, establishes the task relationship path between each collaborative task graph and the public task graph, and obtains the collaborative task relationship graph of each action collaboration unit. Each heavy-duty industrial robot included in the cluster twin model is matched and associated with its own model collaboration points and corresponding collaborative task relationship graph. The working area of each heavy-duty industrial robot in the work twin space is used as the work simulation environment of its respective robot twin. Each robot twin controls a heavy-duty industrial robot in an actual work area, simulating the tasks performed by each robot twin in its own work simulation environment, and thus carrying out distributed simulation control of the robot collaborative cluster in the actual work area.
[0012] In a preferred embodiment, the process of establishing task relationship paths between each collaborative task graph and the common task graph, and obtaining the collaborative task relationship graph of each action collaboration unit, includes: Locate the graph fill elements for the same action parts between each collaborative task graph and the common task graph, and set the position of each graph fill element as a graph connection point. Based on the graph connection points, establish the task relationship path between the collaborative task graph and the common task graph, compile the execution process of the actions corresponding to the graph connection points into action middleware, and call the action middleware to the corresponding collaborative task graph through the task relationship path. When the collaborative task graph executes control for a heavy-duty industrial robot, the motion middleware is used to compile the control data of the execution process of the corresponding motion in the common task graph, and the graph positions that do not call the motion middleware are compiled to complete the operation task for the heavy-duty industrial robot. The action at the latest compiled position in the collaborative task graph is set as the graph filling element of the public task graph. The public task graph is expanded and updated based on the graph filling element. Each collaborative task graph and the public task graph are connected through task relationship paths to obtain their own collaborative task relationship graph.
[0013] The technical effects and advantages provided by the present invention in the above technical solution are as follows: it realizes intelligent decomposition and semantic collaborative modeling of multi-load robot tasks, and supports distributed collaborative pre-playing and verification in the simulation environment by constructing a digital twin cluster and a collaborative task relationship graph, thereby improving the task collaboration efficiency when executing tasks. Attached Figure Description
[0014] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.
[0015] Figure 1This is a flowchart of a task collaboration management method for multi-load industrial robots as described in this invention.
[0016] Figure 2 This is a schematic diagram of the collaborative task relationship graph of the task collaboration management method for multi-load industrial robots described in this invention. Detailed Implementation
[0017] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0018] Example 1, please refer to Figure 1 As shown in the figure, the task collaboration management method for multi-load industrial robots described in this embodiment includes the following steps: Step S1: Establish a robot collaborative cluster for several heavy-duty industrial robots, create a cluster twin model corresponding to the robot collaborative cluster using digital twin technology, and set several model collaboration points based on the distribution location of the heavy-duty industrial robots in the cluster twin model. Step S2: Set up the task and send it to the control terminal that has a communication connection with several heavy-duty industrial robots. In the control terminal, the task is processed into several sub-tasks. The sub-tasks are semantically understood and associated with each other to create a node unit, and the corresponding action collaboration unit of each sub-task is obtained. Step S3: Establish a collaborative task relationship graph between different action collaboration units. Based on the collaborative task relationship graph and the cluster twin model, perform distributed simulation control of the robot collaborative cluster in the operation simulation environment.
[0019] It should be further explained that, in a specific implementation, the process of establishing a robot collaborative cluster for several heavy-duty industrial robots and creating a cluster twin model corresponding to the robot collaborative cluster using digital twin technology includes: Several heavy-duty industrial robots are labeled and denoted as i, then i = 1, 2, 3, ..., n, where n is a natural number greater than 0. Each heavy-duty industrial robot is assigned a file interface, which is used to obtain the relevant data of the heavy-duty industrial robot and generate a digital file based on the relevant data. Digital archives are used to record the status and operation information of heavy-duty industrial robots. The status information includes the status archives of the heavy-duty industrial robot when it is in working state and standby state. The operation information includes the operation records of the heavy-duty industrial robot when it performs different operation procedures. All operation records are integrated to generate the operation archive corresponding to the heavy-duty industrial robot. Status profiles are used to obtain the core motion and pose status, system load and performance status, electrical and thermal status, safety and fault status, and configuration and maintenance status of the heavy-duty industrial robot itself during operation. Among them, the core motion and pose states include joint-level information consisting of joint angles, joint velocities, joint accelerations, and motor current / torque of the robotic arm at different positions on the heavy-duty industrial robot, as well as end effector-level information consisting of the robotic arm's Cartesian space pose, end-effector linear velocity and angular velocity, and end-effector force / torque. System load and performance status are used to characterize the current working intensity and capability boundaries of heavy-duty industrial robots. Specifically, they include the load rate of each axis, which is composed of the percentage of real-time torque and rated torque of the motor corresponding to each joint on the robotic arm, and the percentage of the maximum achievable speed / acceleration of each joint on the robotic arm based on motion parameters, relative to the percentage of the preset maximum rated achievable speed / acceleration of the heavy-duty industrial robot. The electrical and thermal status includes temperature information consisting of the servo motor temperature, reducer / gearbox temperature, and key components of the driver / controller corresponding to the heavy-duty industrial robot, as well as electrical parameters consisting of main circuit voltage / current, brake status, and battery status. Safety and fault status includes fault and alarm codes corresponding to heavy-duty industrial robots. Specifically, the fault and alarm codes are controller diagnostic codes and safety board diagnostic codes. The controller diagnostic codes are used to indicate abnormal problems such as overcurrent, overvoltage, overheating, encoder failure, and communication timeout. The safety board diagnostic codes are used to indicate faults related to safety circuits. Configuration and maintenance status includes the static capability parameters and maintenance-related information of the heavy-duty industrial robot. The static capability parameters include model, rated load, maximum arm span and repeatability accuracy, including the workspace range of the robotic arm, as well as the type and parameters of the end effector represented by the maximum opening and closing degree of the gripper and the vacuum degree of the suction cup. The maintenance-related information includes the total power-on time of the robotic arm, the total running time / number of cycles of each axis, the maintenance timer of key components (such as the reducer) and the lubrication status.
[0020] A communication port is set up at the location of each heavy-duty industrial robot. A corresponding communication interaction layer is created based on the communication port. All communication interaction layers are connected to build a communication interaction network. The heavy-duty industrial robots exchange data based on the communication interaction network, thereby establishing a robot collaborative cluster corresponding to several heavy-duty industrial robots.
[0021] The working areas of several heavy-duty industrial robots are modeled using digital twin technology to generate corresponding working twin spaces. A robot twin is instantiated for each heavy-duty industrial robot using digital twin technology. Based on the actual position coordinates of each heavy-duty industrial robot in the work area, the corresponding robot twin is mapped to the corresponding spatial coordinate position in the work twin space. A twin interaction network is created based on the communication interaction network, and the twin interaction network is deployed to the work twin space. Data interaction between robot twins is carried out based on the twin interaction network, thereby creating a cluster twin model.
[0022] It should be further explained that, in the specific implementation, the process of setting the cluster twin model based on the distribution location of heavy-duty industrial robots as several model collaboration points includes: Obtain the planned operation route for each heavy-duty industrial robot, and record the intersection of the operation routes of different heavy-duty industrial robots as the route interchange point. When different heavy-duty industrial robots are at the same route interchange point, establish a collaborative relationship for the robot twins representing the different heavy-duty industrial robots in the operation twin space.
[0023] Determine whether the collaboration relationship has been successfully established. If so, set the route interchange point where the collaboration relationship has been established as the model collaboration point. Create a corresponding task collaboration space for the two robot twins under the collaboration relationship. In the task collaboration space, the different robot twins will cooperate to process the received task and complete the current task together. If not, then the collaborative relationship will be re-established.
[0024] The process of establishing a collaborative relationship between different robot twins involves: setting a corresponding safe operating radius for each robot twin; obtaining a corresponding safe operating area based on the safe operating radius; using the intersection of the safe operating areas of two robot twins as the collaborative area; obtaining the preset task collaboration procedures for different robot twins within the collaborative area; arranging all the task collaboration procedures based on the preset task form; and generating a work procedure to be executed by different robot twins within the collaborative area. If the work procedure matches all the procedures of a complete work task, the collaborative relationship between different robot twins is considered successfully established; otherwise, no collaborative relationship is formed.
[0025] It should be further explained that, in the specific implementation, the process of setting up a task and sending it to a control terminal connected to several heavy-duty industrial robots, and then processing the task into several sub-tasks within the control terminal, includes: Set up a control terminal, which has several control containers. Each control container is used to establish a transmission path that connects to the heavy-duty industrial robot and executes equipment control. The transmission path is subject to bidirectional security verification. The transmission path that passes the two-way security verification is marked as a two-way interactive path. The transmission path that fails the two-way security verification is driven by security events. Several dangerous nodes on the transmission path are obtained through the security event driving, and each dangerous node is handled abnormally, thereby processing the corresponding transmission path into a two-way interactive path.
[0026] When the heavy-duty industrial robot and the control terminal are in a two-way interactive path, the control terminal is used to establish a communication connection with the heavy-duty industrial robot and then control the corresponding heavy-duty industrial robot. When the heavy-duty industrial robot and the control terminal are not in a two-way interactive path, the control terminal does not control the heavy-duty industrial robot. A job task is set up, which consists of several process steps. These process steps are interdependent in execution order and have independent execution relationships. The control terminal receives the job task and processes it into branch subtasks based on the execution relationships of the process steps.
[0027] Among them, different process steps that are interdependent in execution relationship are each treated as a branch sub-task, and are assigned to heavy-duty industrial robots that perform the work in different time periods based on the execution order. Process steps that are independent in execution relationship are each treated as a branch sub-task, and the same processing time period is set for the independent branch sub-tasks in execution relationship, and the heavy-duty industrial robots in standby state are called to process them. The control of heavy-duty industrial robots within the work area is based on the compilation of corresponding control scripts by the respective robot twins within the work twin space. These control scripts are created based on the relevant data corresponding to the work tasks.
[0028] It should be noted that a control script is a structured, machine-executable set of instructions and data packets. It is a comprehensive program that integrates a series of precise action sequences, collaborative logic, safety constraints, and digital twin mapping relationships for heavy-duty industrial robots. Its goal is to drive the heavy-duty industrial robot in the actual work area and its robot twin in the work twin space to synchronously and precisely complete specified branch sub-tasks while ensuring safety and collaboration. Specifically, the control script includes script header metadata and an identity identifier, used to define the basic information and execution context of the control script. The specific content of the script header metadata and identity identifier includes a global task ID and a subtask ID: associated with the upper-level work task and its subtask ID. The decomposed sub-tasks include: a unique identifier for the target robot indicating which heavy-duty industrial robot will execute the control script; security level and permissions defining the minimum security permissions (e.g., operating mode, area permissions) required for script execution; and an associated twin identifier specifying the particular robot twin bound to the current heavy-duty industrial robot in the work twin space. The control script specifically includes the precise motion path of the heavy-duty industrial robot's end effector or key joints in space. For example, a sequence of ordered path points on the corresponding mobile work path of the heavy-duty industrial robot, where each path point contains pose data: the target position (X, ) in a pre-set workpiece coordinate system. Y, Z) and robot posture, motion type: straight line segments or curve segments formed between different path points, speed and acceleration percentage: percentage relative to the robot's preset maximum moving speed, acceleration or standard value. The control script also specifically includes the motion trajectory executed by the robotic arm of the heavy-duty industrial robot, or the process operation to be performed when it reaches a specific coordinate point. For example, the process operation corresponding to the end effector: for the fixture, set the pressure value and opening of the fixture; for the welding gun, set the working current, working voltage and welding speed of the welding gun; for the glue gun, set the spraying flow rate of the glue gun; the process operation corresponding to the external equipment: when the pallet is in place, wait for the sensor to send a control signal and then control the external cylinder or power tool.
[0029] The action sequence is converted into executable script instructions. Based on the structured specifications of control scripts, and combined with the motion parameters of heavy-duty industrial robots, process operation requirements, and digital twin mapping relationships, the transformation from abstract action logic to machine-recognizable instructions is achieved, as detailed below: The action collaboration units corresponding to the branch subtasks are decomposed, and the core elements of each action are extracted to form a standardized parameter set. The standardized parameter set includes basic identity parameters, motion trajectory parameters, process operation parameters, and twin synchronization parameters. Basic identity parameters are used to ensure the legitimacy and relevance of script execution; This includes the associated global task ID, subtask ID, target robot unique identifier, and bound robot twin identifier, which are filled in according to the script header metadata format, while matching the preset security level and permission parameters; Analyze the motion path of the end effector or key joint of the robotic arm in the motion sequence, extract the coordinate information and attitude data of the path point sequence, and generate the corresponding motion trajectory parameters; The coordinate information needs to be converted to standard values in the workpiece coordinate system, and the posture data needs to correspond to the robot joint angle combination or Euler angle parameters; the motion type (straight line segment / curved segment) between path points should be distinguished, and the velocity and acceleration percentage (relative to the robot's maximum rated value) of each segment should be marked to form a motion trajectory parameter table. Extract the end effector operation instructions and external device linkage instructions contained in the action sequence. The end effector operation requires the clear identification of the equipment type and corresponding process parameters, and then generate the process operation parameters, such as the pressure threshold and opening / closing range of the fixture, the working current, voltage, and welding speed of the welding torch, and the spraying flow rate and spraying time of the glue gun. The external device linkage requires the clear identification of the triggering conditions (such as sensor signals and time delays), the execution object (such as cylinders and power tools), and the action instructions (such as extension / retraction, start / stop).
[0030] The correlation parameters between the robot's actual actions and the mapping between the twin are extracted as twin synchronization parameters, including the timestamp synchronization rules for action execution, the frequency of state data feedback, and the triggering conditions for twin posture updates, to ensure that the states of the physical robot and the digital twin are consistent in real time during script execution; Based on the robot instruction set syntax specification supported by the control terminal, the extracted standardized parameters are converted into machine-executable script instructions, specifically including head instruction encoding, motion instruction encoding, process instruction encoding, and twin synchronization instruction encoding. The header instruction encoding follows the script header metadata format, writing identity parameters and security parameters as fixed fields at the beginning of the script, with fields separated by delimiters. It also adds extended parameters such as script version number and execution validity period to facilitate script management and traceability. Motion command encoding is based on motion trajectory parameters and calls the corresponding functions in the robot motion control command library, such as the linear motion command LINE_MOVE and the curvilinear motion command ARC_MOVE. Parameters such as path point coordinates, attitude data, motion speed, and acceleration are used as function input parameters and encoded in the format of command keyword + parameter list. For continuous path motion, a path smoothing transition command is added to avoid the robotic arm from stopping suddenly or shaking at path points. The process instruction encoding is based on process operation parameters, calling end effector control instructions and external device linkage instructions, such as fixture control instruction CLAMP_CONTROL (pressure value, opening degree), welding torch control instruction WELD_CONTROL (current, voltage, speed), and external device trigger instruction DEVICE_TRIGGER (device ID, trigger condition, action instruction). For process operations with a sequential execution order, instruction execution delay parameters or condition judgment statements are added to ensure that the operation process meets the process requirements. Twin synchronization instruction encoding is used to add state synchronization instructions for the digital twin corresponding to the physical robot, such as TWIN_SYNC(timestamp, state data type, feedback frequency), which specifies the state data content (such as joint angle, load rate, temperature) and feedback period that the physical robot feeds back to the digital twin. At the same time, twin posture correction instructions are added. When the posture deviation between the physical robot and the twin exceeds the threshold, the twin posture adjustment is automatically triggered. Perform logical validation on the encoded script instructions, specifically including: Syntax check: Call the command syntax checker in the control terminal to verify whether the script command format conforms to the standard, whether the parameter types match, whether there are errors in function calls, mark commands with syntax errors and automatically correct them; Logical conflict verification: Analyze the execution order of motion instructions and process instructions in the script to check for problems such as overlapping motion trajectories and conflicting process operation timing. For example, the robotic arm may execute a process operation before reaching the target position, or the motion trajectories of two robots may interfere with each other in the collaborative area. For conflicting instructions, adjust the instruction execution order or motion parameters based on the simulation results of the collaborative task relationship graph and the cluster twin model. Safety constraint verification: Combining the robot's safe operating radius and collaborative area range, verify whether the path points in the motion command exceed the safe operating area, and whether the parameters in the process command meet the equipment safety thresholds (such as motor torque not exceeding the rated value, temperature not exceeding the warning value); for commands that violate safety constraints, automatically add safety restrictions or terminate the command; According to the preset script encapsulation format, the header instructions, motion instructions, process instructions, and twin synchronization instructions are integrated into a complete script file. A file verification code is added to ensure that the script is not tampered with during transmission. Based on a two-way interactive path, the encapsulated script file is sent to the controller of the target heavy-duty industrial robot and simultaneously sent to the corresponding robot twin. Before the script is executed, the control terminal verifies the reception status of the physical robot and the twin through the communication interaction network and the twin interaction network, respectively, to ensure that both successfully load the script instructions.
[0031] The content of constructing a two-way interaction path is as follows: Within the control terminal, each heavy-duty industrial robot to be connected is configured with an independent control container. The control container is a software execution environment with resource isolation characteristics, with a built-in dedicated communication protocol stack and security policy engine. The control terminal initiates a connection request to the controller of the target heavy-duty industrial robot. Each control container independently maintains a logical transmission path. The transmission path includes physical network links, communication protocol sessions, and logical endpoints between the terminal and the controller. After the transmission path is established, two-way security verification is immediately initiated between the control terminal and the controller of the heavy-duty industrial robot, including first-direction verification and second-direction verification. The first direction verification is the verification of the controller of the heavy-duty industrial robot by the control terminal. The control terminal sends a challenge request containing a random number and a timestamp to the controller. The controller signs the challenge request using its device private key and returns the signature result, digital certificate and system integrity hash value. The control terminal verifies the validity of the signature and checks whether the integrity hash value is consistent with the pre-stored trusted benchmark value. The second direction verification is the verification of the control terminal by the controller of the heavy-duty industrial robot. The controller requests the control terminal to provide control authorization credentials and terminal security status proof. The control terminal returns an electronic license issued by the superior system and its current security status information. The controller verifies the validity of the electronic license and evaluates the terminal security status based on preset strategies.
[0032] If both the first and second direction verifications are successful, the control terminal marks the corresponding transmission path as a bidirectional interactive path, and the path enters an available state, allowing subsequent bidirectional secure transmission of commands and data; if verification in either direction fails, a security event-driven process is triggered, the transmission path is marked as unverified, and any control commands are prohibited from passing through.
[0033] For transmission paths where two-way security authentication fails, the following handling procedure shall be performed: Deploy a security event-driven engine corresponding to the transmission path. The security event-driven engine diagnoses along the transmission path based on the verification failure type and locates the specific dangerous nodes that caused the failure. Dangerous nodes include identity nodes, integrity nodes, or network nodes. Execute the corresponding anomaly handling plan according to the type of dangerous node. For identity or integrity nodes, perform credential update, baseline repair or device isolation. For network nodes, initiate network access control or deep inspection. After the handling is completed, re-execute the two-way security verification for the corresponding transmission path.
[0034] For transmission paths that have successfully undergone bidirectional security verification after anomaly handling, their status is updated to bidirectional interactive paths. For established bidirectional interactive paths, the control terminal performs periodic or trigger-based lightweight reverification. If a path anomaly is detected or a security threat intelligence is received during operation, the path is immediately downgraded to an unverified state, and the security event-driven process is retried. Through the above steps, a bidirectional, reliable, secure, and controllable communication channel is built and maintained between the control terminal and the heavy-duty industrial robot, laying a secure communication foundation for the subsequent issuance of precise control commands based on digital twins.
[0035] It should be further explained that, in the specific implementation, the process of semantically understanding the branch subtasks and creating a node unit for each subtask to obtain the corresponding action collaboration unit includes: A bag-of-words model is constructed based on text analysis technology. The bag-of-words model is used to perform semantic understanding of the task content of the branch sub-tasks and filter out several task keywords corresponding to the branch sub-tasks. Each task keyword corresponds to one action performed by the heavy-duty industrial robot. Each unique task keyword is constructed as an action matching node. All action matching nodes corresponding to the same branch subtask are aggregated, and node units corresponding to the corresponding branch subtask are created. Each action matching node is assigned a unique node identifier within the node unit, which is used as the identity authentication identifier for each action matching node.
[0036] Associate each branch subtask with the corresponding node unit of the branch subtask; Obtain historical task data of heavy-duty industrial robots performing tasks, process the historical task data based on big data technology, and obtain the matching relationship between each actual action performed by the heavy-duty industrial robot and the instruction keywords corresponding to the issued instructions. The matching relationship is denoted as: instruction keyword - execution action. A database is set to store all matching relationships. The database containing all matching relationships is marked as the action library. The node unit corresponding to each branch subtask is imported into the action library. Then, based on the task keyword of the included action matching node, the node unit selects all execution actions corresponding to instruction keywords that match the task keyword from the action library. All the obtained execution actions are integrated as the action collaboration unit of the corresponding branch subtask. The above operation is repeated until the action collaboration unit corresponding to each branch subtask is obtained.
[0037] It should be further explained that, in the specific implementation, the process of establishing a collaborative task relationship graph between different action collaboration units, and based on the collaborative task relationship graph and the cluster twin model, performing distributed simulation control of the robot collaborative cluster in the work simulation environment includes: Select one of several action collaboration units as the initial public graph filling object, execute all actions of the action collaboration unit corresponding to the public graph to control the heavy-duty industrial robot, and use each action as a graph filling element of the graph filling object in turn. When the initial public graph has completed filling all graph filling elements under the entire graph filling object, the initial public graph is converted into a public task graph. The public task graph is used to represent the execution process of all actions of the corresponding action collaboration unit.
[0038] Repeat the steps above to establish a common task graph, and establish a corresponding collaborative task graph for each other action collaboration unit. Locate the graph filling element corresponding to the same action part between each collaborative task graph and the common task graph, and set the position of each graph filling element obtained by the location as a graph connection point. Based on the graph connectors, establish task relationship paths between the collaborative task graph and the common task graph, compile the execution process of the actions corresponding to the graph connectors into action middleware, and call the action middleware to the corresponding collaborative task graph through the task relationship paths.
[0039] When the collaborative task graph needs to perform control of a heavy-duty industrial robot, the action middleware is used to correspond to the control data of the execution process of the corresponding action in the common task graph, and the graph positions in the collaborative task graph that do not call the action middleware are compiled, thereby completing the work task of the collaborative task graph for the heavy-duty industrial robot. The action corresponding to the latest compiled graph position in the collaborative task graph is set as the latest graph filling element in the public task graph. The public task graph is then expanded and updated based on the graph filling element. Each collaborative task graph and the public task graph are connected through task relationship paths to obtain their own collaborative task relationship graph.
[0040] For a structural diagram of the collaborative task relationship graph, please refer to [link / reference needed]. Figure 2 As shown; Each heavy-duty industrial robot included in the cluster twin model is matched and associated with its own model collaboration points and corresponding collaborative task relationship graph. The working area of each heavy-duty industrial robot in the work twin space is used as the work simulation environment of its respective robot twin. Each robot twin controls a heavy-duty industrial robot in an actual work area, simulating the tasks performed by each robot twin in its own work simulation environment. This leads to the distributed simulation control of the robot collaborative cluster composed of all heavy-duty industrial robots in the actual work area.
[0041] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
Claims
1. A task collaboration management method for multiple load industrial robots, characterized by, Comprise the following steps: Step S1: Establish a robot collaboration cluster for several heavy-load industrial robots, create a cluster twin model corresponding to the robot collaboration cluster through digital twin technology, and set the cluster twin model based on the distribution position of the heavy-load industrial robots as several model collaboration points; Step S2: Set a job task and issue it to a control terminal connected with several heavy-load industrial robots, process the job task into several branch sub-tasks in the control terminal, perform semantic understanding on the branch sub-tasks and associate to create a node unit for each, and obtain the action collaboration unit corresponding to each branch sub-task; Step S3: Establish a collaborative task relationship graph between different action collaboration units, and based on the collaborative task relationship graph and the cluster twin model, perform distributed simulation control on the robot collaboration cluster in the job simulation environment.
2. The task coordination management method for multiple load industrial robots according to claim 1, wherein, The process of establishing a robot collaboration cluster for several heavy-load industrial robots and creating a cluster twin model corresponding to the robot collaboration cluster through digital twin technology comprises: Assign an archive interface to each heavy-load industrial robot for obtaining relevant data of the heavy-load industrial robot to generate a digital archive, which is used to record the state information and operation information of the heavy-load industrial robot; Set a communication port at the position of each heavy-load industrial robot, create a respective communication interaction layer based on the communication port, connect all the communication interaction layers to build a communication interaction network, and perform data exchange based on the communication interaction network to establish a robot collaboration cluster corresponding to the several heavy-load industrial robots; Model the several heavy-load industrial robots and the working area of the heavy-load industrial robots through digital twin technology to generate a corresponding job twin space and a robot twin body, map the corresponding robot twin body to the corresponding spatial coordinate position of the job twin space based on the actual position coordinates of the heavy-load industrial robots in the working area, and create a cluster twin model.
3. The task coordination management method for multiple load industrial robots according to claim 2, wherein, The process of setting the cluster twin model based on the distribution position of the heavy-load industrial robots as several model collaboration points comprises: Obtain the working route of each heavy-load industrial robot, record the intersection point of different working routes as a route interworking point, and build a collaboration relationship between different robot twins in the twin space at the same route interworking point; Determine whether the collaboration relationship is successfully built, if yes, set the route interworking point where the collaboration relationship is successfully built as a model collaboration point; if not, rebuild the collaboration relationship.
4. The task coordination management method for multiple load industrial robots according to claim 3, wherein, The process of building a collaboration relationship between different robot twins comprises: Set a safe working radius for each robot twin, obtain a respective safe working area based on the safe working radius, and take the intersection area corresponding to the safe working areas of two robot twins as a collaboration area; In the collaboration area, obtain the preset task collaboration procedures of different robot twins, arrange all the task collaboration procedures based on the preset task form, generate a working procedure corresponding to the execution of different robot twins in the collaboration area, and if the working procedure meets all the procedures of a complete job task, the collaboration relationship is successfully built, otherwise, the collaboration relationship is not formed.
5. The task coordination management method for multiple load industrial robots according to claim 4, wherein, The process of setting a work task and issuing it to a control terminal connected with several heavy-load industrial robots includes: setting a control terminal including several control containers, each of which is used to establish a transmission path for controlling the heavy-load industrial robots, and performing bidirectional security verification on the transmission path; marking the transmission path that passes the bidirectional security verification as a bidirectional interaction path; when in the bidirectional interaction path, the control terminal is connected with and controls the heavy-load industrial robots, and when not in the bidirectional interaction path, the control terminal does not control the heavy-load industrial robots; setting a work task, which is composed of several process steps that have a mutual dependence in execution sequence and an independent execution relationship, and receiving the work task by the control terminal, and processing the work task into branch sub-tasks based on the execution relationship of the process steps.
6. The task coordination management method for multiple load industrial robots according to claim 5, wherein, The process of performing semantic understanding on the branch sub-tasks and creating a node unit for each branch sub-task includes: building a bag-of-words model to perform semantic understanding on the task content of the branch sub-tasks, and selecting several task keywords of the branch sub-tasks, each of which corresponds to an action performed by the heavy-load industrial robots; building each non-repeated task keyword into an action matching node, aggregating all action matching nodes of the same branch sub-task, creating a node unit of the branch sub-task, and associating each branch sub-task with the node unit corresponding to the branch sub-task; obtaining the matching relationship between each actual execution action of the heavy-load industrial robots and the instruction keyword corresponding to the issued instruction, setting a database to store all matching relationships, and identifying the database as an action library; importing the node unit of each branch sub-task into the action library, selecting all execution actions corresponding to the instruction keywords that match the task keywords from the action library, and integrating all execution actions as the action cooperation unit of the corresponding branch sub-task.
7. The task coordination management method for multiple load industrial robots according to claim 6, wherein, The process of establishing a collaborative task relationship graph between different action cooperation units and performing distributed simulation control on the robot cooperation cluster in the work simulation environment based on the collaborative task relationship graph and the cluster twin model includes: selecting an action cooperation unit as an initial common graph filling object, and executing all actions of the selected action cooperation unit to control the heavy-load industrial robots, each action being a graph filling element of the graph filling object; when the initial common graph is filled with all graph filling elements, it is converted into a common task graph, each action cooperation unit establishes its own collaborative task graph, a task relationship path is established between each collaborative task graph and the common task graph, and the collaborative task relationship graph of each action cooperation unit is obtained; based on the model cooperation point of each heavy-load industrial robot included in the cluster twin model and the corresponding collaborative task relationship graph, each heavy-load industrial robot is matched and associated, and the work area of each heavy-load industrial robot in the work twin space is used as the work simulation environment of the robot twin body. Each robot twin controls a heavy industrial robot in an actual work area, executes simulation of each robot twin's task work in a respective work simulation environment, and accordingly performs distributed simulation control of the robot collaboration cluster in the actual work area.
8. The task coordination management method for multiple load industrial robots according to claim 7, wherein, The process of establishing a task relationship path between each collaboration task graph and the common task graph, and obtaining a collaborative task relationship graph of each action collaboration unit, includes: Positioning the graph filling elements of the same action part between each collaboration task graph and the common task graph, and setting the positions of each graph filling element as graph connection points; Based on the graph connection points, a task relationship path is established between the collaboration task graph and the common task graph, the execution process of the corresponding action of the graph connection point is compiled into an action middleware, and the action middleware is called to the corresponding collaboration task graph through the task relationship path; When the collaboration task graph executes the control of the heavy industrial robot, the control data of the execution process of the corresponding action in the common task graph is used to compile the graph positions of the action middleware, and the work task of the heavy industrial robot is completed; The action at the latest compiled graph position in the collaboration task graph is set as the graph filling element of the common task graph, the common task graph is extended and updated based on the graph filling element, and the respective collaborative task relationship graph is constructed between each collaboration task graph and the common task graph through the task relationship path.