A multi-sensor-based intelligent supervision method and system for welding robots
By using multi-sensor real-time monitoring and dynamic obstacle avoidance trajectory planning, the problems of collision and waiting in multi-robot collaborative welding have been solved, enabling safe and efficient collaborative operation and improving the overall efficiency and stability of the production line.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA ENTERPRISE DYNAMIC ROBOT TECH JIANGSU CO LTD
- Filing Date
- 2026-02-27
- Publication Date
- 2026-06-12
AI Technical Summary
In multi-robot collaborative welding operations, traditional fixed trajectory planning is prone to robot arm collisions or task stagnation, resulting in decreased production efficiency and a lack of intelligent monitoring methods for real-time perception and dynamic coordination.
By collecting real-time operational data of the welding robot through multiple sensors and combining it with historical monitoring data to identify trajectory intersection risks, dynamic obstacle avoidance operation trajectories are generated and coordination commands are triggered, thereby realizing real-time optimization and dynamic adjustment of the robot's operation path.
It improves the safety and production efficiency of multi-robot collaborative operations, avoids collision and waiting issues, and enhances the system's adaptability and long-term operational stability.
Smart Images
Figure CN122185166A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of welding robot technology, specifically to a multi-sensor-based intelligent monitoring method and system for welding robots. Background Technology
[0002] In automated welding production lines, multi-robot collaborative operation is a crucial means to improve production efficiency and task parallelism. Traditional welding robot operation planning typically relies on fixed planned trajectories, meaning the welding robots perform welding operations at independent workstations according to preset paths. This method can ensure the basic operation of the production process when the welding task is simple and the robot workspaces do not overlap. In practical applications, to improve the overall production efficiency of the assembly line, multi-robot collaborative welding operations are commonly adopted. However, when multiple robots collaborate in welding operations, the movement trajectories of the robots often inevitably overlap. In this case, following fixed planned trajectories can easily lead to physical collisions between robotic arms, or to a state of mutual waiting and task stagnation to avoid collisions, resulting in a decrease in overall production efficiency. Therefore, there is an urgent need for an intelligent monitoring method and system capable of real-time perception and dynamic coordination to achieve safe and efficient collaborative operation of multiple welding robots in a shared workspace. Summary of the Invention
[0003] The purpose of this invention is to provide a method and system for intelligent supervision of welding robots based on multiple sensors, so as to solve the problems raised in the prior art.
[0004] To achieve the above objectives, the present invention provides the following technical solution: an intelligent monitoring method for welding robots based on multiple sensors, the intelligent monitoring method comprising: Step S100: Obtain historical operation monitoring data of the automated welding production line and capture the operation information of each welding robot in the production line; transmit the operation information to the welding supervision terminal in a unified manner; wherein, the automated welding production line is a welding production line with multiple robots working collaboratively; the welding supervision terminal integrates and processes the welding robot operation information of the production line through sensors and captures target welding robots with the risk of trajectory intersection. Step S200: Obtain the work area layout plan of the automated welding production line, mark the target welding robot in the work area layout plan, and extract feature data of the target welding robot; Step S300: Based on the feature data of each target welding robot, perform correlation judgment on the operation trajectory between every two target welding robots to determine the trajectory associated robot of each target welding robot; Step S400: Based on the distribution of the target welding robots associated with their trajectories, generate several segments of dynamic obstacle avoidance trajectories in the work area layout diagram, and store the dynamic obstacle avoidance trajectories. Step S500: Accumulate the welding operation conditions corresponding to each target welding robot; when the accumulated value of the operation condition corresponding to a target welding robot reaches the collision warning threshold of the target welding robot, generate the dynamic coordination command of the target welding robot; at the same time, extract the dynamic obstacle avoidance operation trajectory with the target welding robot as the trajectory starting point. Step S600: Simultaneously transmit the dynamic coordination command and the dynamic obstacle avoidance operation trajectory to the operation terminal of the corresponding target welding robot, and control and assist each target welding robot to complete the safe and collaborative welding operation.
[0005] Furthermore, step S100 includes: Step S101: Capture the equipment parameters and operating data of each welding robot in the automated welding production line through sensors, lock the welding operation trajectory of each welding robot according to the equipment parameters of each welding robot, and capture the working space range of each welding robot in the process of executing the corresponding welding operation trajectory. The working space range is obtained by fitting a three-dimensional space model through the coordinates of the welding robot's motion limit position collected by the sensors. Step S102: If the workspace of a welding robot overlaps with that of other welding robots, the welding robot is designated as the target welding robot. The overlapping area is the work area where trajectory intersections are likely to occur during multi-robot collaborative operations. The spatial area and position coordinates of the overlapping area are obtained by intersecting the three-dimensional spatial models of the two robots. The probability of trajectory intersection between each target welding robot and other robots during the execution of the corresponding welding operation trajectory is obtained. The trajectory intersection probability is obtained by dividing the number of trajectory intersections in the historical operation cycle by the total number of operations in the historical operation cycle, multiplying the spatial area of the overlapping area by the total area of the target robot's workspace. The number of trajectory intersections in the historical operation cycle is obtained by statistically analyzing the robot trajectory coordinate overlap records in the historical operation monitoring data collected by the sensors, and the total number of operations in the historical operation cycle is obtained by statistically analyzing the start and stop records of the welding robot's operations.
[0006] Furthermore, based on the trajectory intersection probability of each target welding robot, a collision warning threshold is set for each target welding robot to trigger the generation of dynamic coordination commands. The collision warning threshold is the cumulative critical value of the working conditions of each target welding robot, and is calculated as follows: the area of the overlapping area is divided by the total area of the target robot's working space to obtain the area ratio. The area ratio, trajectory intersection probability, and historical collision accident frequency are then weighted and summed to obtain the collision warning value. The weighting coefficient is preset according to the process complexity and safety level of the welding operation on the production line. The historical collision accident frequency is the number of times the target robot has physically collided with other robots in the historical operation cycle, which is obtained by statistically analyzing the fault alarm records in the historical operation monitoring data.
[0007] Furthermore, step S200 includes: The collision warning threshold for each target welding robot when triggering the dynamic coordination command is obtained, and this collision warning threshold is used as the first feature data P1 of each target welding robot. The response time of each target welding robot for each completion of the work trajectory adjustment is obtained and the average response time is calculated. This average response time is used as the second feature data P2 of each target welding robot. The trajectory adjustment response time is the time interval from sending the trajectory adjustment command from the welding monitoring terminal to the start of the welding robot's trajectory adjustment action, which is obtained by statistically analyzing the difference between the command sending time and the robot action triggering time collected by the sensor. The operation time required for each target welding robot to complete one corresponding welding operation is obtained and the average operation time is calculated. This average operation time is used as the third feature data P3 of each target welding robot. The operation time of the welding operation is the time interval from the start of the welding operation by the welding robot to the completion of the welding process, which is obtained by statistically analyzing the difference between the start and stop times of the welding operation collected by the sensor.
[0008] Furthermore, step S300, which involves determining the correlation between the work trajectories of every two target welding robots, includes: Step S301: Query the work area information of the target welding robot by dividing the production line workstation data. If the two target welding robots are not in the same work area, it is determined that there is no work trajectory association between the two target welding robots. If the two target welding robots are in the same work area and complete different welding processes in the same work area, feature data is extracted for the two target welding robots. Step S302: Based on the welding sequence of the two target welding robots in the same work area, the two target welding robots are designated as target welding robot X1 and target welding robot X2. The welding process of target welding robot X1 is located before the welding process of target welding robot X2. The first feature data P(X1,1), second feature data P(X1,2), and third feature data P(X1,3) of target welding robot X1 and the first feature data P(X2,1), second feature data P(X2,2), and third feature data P(X2,3) of target welding robot X2 are obtained respectively. The motion data of the welding robots is collected by sensors, and the average trajectory adjustment speed V of the welding robots is obtained by statistical analysis based on the motion data. The calculation method is the sum of the movement distance of each trajectory adjustment divided by the sum of the movement time of each trajectory adjustment. The movement distance and movement time of the trajectory adjustment are obtained by statistical analysis of the changes in the trajectory coordinates and time of the welding robots collected by sensors. The shortest working distance S between target welding robot X1 and target welding robot X2 is calculated by using the coordinate data of the work area layout plan. Step S303: In the historical operation monitoring data of the automated welding production line, when the cumulative value of the working condition of the target welding robot X1 reaches the first feature data P(X1,1), calculate the average cumulative value of the working condition of the corresponding target welding robot X2 P(X2,u). The average cumulative value of the working condition P(X2,u) is calculated as follows: the sum of the cumulative values of the working conditions of X2 when the cumulative value of X1 reaches P(X1,1) divided by the total number of times the statistics are performed. If P(X2,u)≥P(X2,1), it is determined that the target welding robot X1 and the target welding robot X2 satisfy the operation trajectory association, and the target welding robot X2 is the trajectory association robot of the target welding robot X1. If P(X2,u) < P(X2,1), using the second feature data P(X1,2), the shortest working distance S, and the average trajectory adjustment speed V, calculate the shortest time for the target welding robot X1 to move to a non-crossing risk working position after completing one trajectory adjustment. If, after completing one work trajectory adjustment, the target welding robot X1 moves to a work position with no cross-risk within the shortest time T, predict the cumulative work condition value P(X2,r) corresponding to the target welding robot X2, where, ,in, Let P(X2,r) represent the number of welding operations that target welding robot X2 can complete within the shortest time T. If P(X2,r)≥P(X2,1), then it is determined that target welding robot X1 and target welding robot X2 satisfy the operation trajectory association, and target welding robot X2 is the trajectory-associated robot of target welding robot X1.
[0009] Furthermore, step S400 includes: Step S401: Obtain the trajectory-associated robots for each target welding robot; in the work area layout plan, calculate the shortest movement path between each target welding robot and its corresponding trajectory-associated robots, and mark the other workstations traversed by the shortest movement path; if the number of other workstations is less than the workstation number threshold, then, taking each target welding robot as the trajectory starting point, generate a dynamic obstacle avoidance operation trajectory in the work area layout plan that covers the corresponding trajectory-associated robots and other workstations, and store the dynamic obstacle avoidance operation trajectory; wherein, the workstation number threshold is preset according to the complexity of the production line operation, the shortest connection path is calculated using the Dijkstra algorithm, and the movement speed limit of each segment is matched according to the average trajectory adjustment speed V; Step S402: If the number of other workstations is greater than the workstation number threshold, then taking each target welding robot as the trajectory starting point, in the work area layout diagram, combining the shortest working movement distance S and the average trajectory adjustment speed V of the welding robot, the shortest dynamic obstacle avoidance operation trajectory from each target welding robot to each trajectory-related robot is generated, and the shortest dynamic obstacle avoidance operation trajectory is stored; wherein, the shortest dynamic obstacle avoidance operation trajectory is obtained by calculating the shortest path between two points combined with the robot motion obstacle avoidance constraint, the obstacle avoidance constraint is the safe distance between the robot arm and surrounding equipment when it moves, and this safe distance is preset by the robot equipment parameters and the workstation layout size.
[0010] Furthermore, when accumulating the real-time welding operation status of each target welding robot, the cumulative value of the real-time welding operation status of each target welding robot will be cleared to zero after the target welding robot completes the obstacle avoidance collaborative operation according to the dynamic obstacle avoidance operation trajectory. The above steps match the robot's accumulated risk status with the actual operational safety status, thus resetting the risk status from a data perspective.
[0011] Furthermore, in order to better implement the above method, a multi-sensor-based intelligent monitoring system for welding robots is also provided. This intelligent monitoring system includes: a risk identification module, a feature extraction module, an associated trajectory judgment module, a dynamic obstacle avoidance trajectory generation module, an instruction generation module, and a transmission control module. The risk identification module is used to acquire historical operation monitoring data of the automated welding production line, capture the operation information of each welding robot, integrate and process the data collected by the sensors, and identify target welding robots with trajectory intersection risks based on the collected data. The feature extraction module is used to obtain a planar layout of the work area, annotate the target welding robot, and extract the feature data of the target welding robot. The associated trajectory judgment module is used to perform correlation analysis on the operation trajectories between every two target welding robots based on the feature data of the target welding robots, and to determine the associated robots of each target welding robot's trajectory. The dynamic obstacle avoidance trajectory generation module is used to generate and store several segments of dynamic obstacle avoidance operation trajectories in the work area layout map based on the distribution of trajectory-related robots. The instruction generation module is used to accumulate the welding operation conditions of the target welding robot. When the accumulated value reaches the collision warning threshold, it generates dynamic coordination instructions and extracts the corresponding dynamic obstacle avoidance operation trajectory. The transmission control module is used to transmit dynamic coordination commands and dynamic obstacle avoidance operation trajectories to the corresponding target welding robot's operating terminal, assisting it in completing safe and collaborative welding operations.
[0012] Furthermore, the risk identification module includes: a sensor data acquisition unit and a risk target identification unit; The sensor data acquisition unit is used to collect the equipment parameters and operating data of each welding robot in the automated welding production line in real time through sensors, lock the welding operation trajectory of each welding robot and capture its working space range during the operation. The risk target identification unit is used to determine whether there is an overlapping area in the workspace of each welding robot. If there is, the robot is marked as the target welding robot, and the probability of its trajectory intersecting with other robots during operation is calculated.
[0013] Furthermore, the correlation trajectory judgment module includes: a work area query unit and a correlation analysis unit; The work area query unit is used to query the work area information of the target welding robot based on the production line workstation division data, and to determine whether two target welding robots are in the same work area. The correlation analysis unit is used to extract feature data of target welding robots in the same work area, and calculate and determine whether there is a correlation between the two welding robots by combining the welding process sequence, the shortest working distance, and the average trajectory adjustment speed, and then identify the trajectory-correlated robot.
[0014] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention collects the operating data of welding robots in real time through multiple sensors, identifies target welding robots with potential trajectory intersection risks by combining historical monitoring data, and makes judgments on the correlation of operation trajectories based on feature data. This enables accurate identification and early warning of potential collision risks in multi-robot collaborative operations, avoids collision or mutual waiting problems caused by traditional fixed trajectory planning, and significantly improves the operational safety of welding production lines.
[0015] 2. This invention generates and stores dynamic obstacle avoidance operation trajectories. Based on the distribution of trajectory-associated robots and parameters such as the shortest operation distance and average trajectory adjustment speed, it dynamically generates obstacle avoidance trajectories suitable for multi-robot collaborative operations. When the cumulative value of the operation condition reaches the collision warning threshold, it triggers a dynamic coordination command, thereby realizing real-time optimization and dynamic adjustment of the welding robot's operation path and effectively improving the production efficiency of multi-robot collaborative operations.
[0016] 3. This invention sets a collision warning threshold and accumulates welding operation conditions. When the robot approaches a risky state, it generates coordination instructions and obstacle avoidance trajectories in a timely manner. After completing the obstacle avoidance operation, the accumulated value of the operation conditions is cleared to zero, forming a dynamic reset mechanism for the risky state. This avoids the continuous accumulation of risks, improves the system's adaptability and long-term operational stability, and provides an efficient and intelligent monitoring method for automated welding production lines. Attached Figure Description
[0017] Figure 1 This is a schematic diagram of the method flow of a welding robot intelligent monitoring method and system based on multiple sensors according to the present invention; Figure 2 This is a schematic diagram illustrating an application scenario of the intelligent monitoring method and system for welding robots based on multiple sensors according to the present invention. Detailed Implementation
[0018] 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.
[0019] Example: Figure 1-2 As shown, the present invention provides a technical solution: an intelligent monitoring method for welding robots based on multiple sensors. This intelligent monitoring method includes: Step S100: Obtain historical operation monitoring data of the automated welding production line and capture the operation information of each welding robot in the production line; transmit the operation information to the welding supervision terminal in a unified manner; the welding supervision terminal integrates and processes the welding robot operation information of the production line through sensors and captures target welding robots with the risk of trajectory intersection. Step S200: Obtain the work area layout plan of the automated welding production line, mark the target welding robot in the work area layout plan, and extract feature data of the target welding robot; Step S300: Based on the feature data of each target welding robot, perform correlation judgment on the operation trajectory between every two target welding robots to determine the trajectory associated robot of each target welding robot; Step S400: Based on the distribution of the target welding robots associated with their trajectories, generate several segments of dynamic obstacle avoidance trajectories in the work area layout diagram, and store the dynamic obstacle avoidance trajectories. Step S500: Accumulate the welding operation conditions corresponding to each target welding robot; when the accumulated value of the operation condition corresponding to a target welding robot reaches the collision warning threshold of the target welding robot, generate the dynamic coordination command of the target welding robot; at the same time, extract the dynamic obstacle avoidance operation trajectory with the target welding robot as the trajectory starting point. Step S600: Simultaneously transmit the dynamic coordination command and the dynamic obstacle avoidance operation trajectory to the operation terminal of the corresponding target welding robot to control and assist each target welding robot in completing safe and collaborative welding operations; Step S100 includes: Step S101: Capture the equipment parameters and operating data of each welding robot in the automated welding production line through sensors, lock the welding operation trajectory of each welding robot according to the equipment parameters of each welding robot, and capture the working space range of each welding robot in the process of executing the corresponding welding operation trajectory. Step S102: If the workspace of a welding robot overlaps with the workspace of other welding robots, the welding robot is designated as the target welding robot; wherein, the overlapping area is the work area where trajectory intersection is likely to occur when multiple robots work together; the probability of trajectory intersection between each target welding robot and other robots during the execution of the corresponding welding operation trajectory is obtained respectively. Among them, based on the trajectory intersection probability of each target welding robot, a collision warning threshold is set for each target welding robot to trigger the generation of dynamic coordination commands; In an embodiment of the present invention, a multi-sensor system deployed on an automated welding production line collects the operating data of each welding robot in real time. The sensor network transmits the collected equipment parameters, operating trajectories, workspace ranges, and other information to the welding monitoring terminal. The terminal system identifies welding robots with overlapping workspaces through data fusion and spatial modeling, and marks them as target welding robots. At the same time, the system calculates the trajectory intersection probability of each target welding robot based on historical data, providing a basis for subsequent dynamic coordination. Step S200 includes: The collision warning threshold of each target welding robot when the dynamic coordination command is triggered is obtained, and the collision warning threshold is used as the first feature data P1 of each target welding robot; the response time of each target welding robot in each completion of the work trajectory adjustment is obtained and the average response time is calculated, and the average response time is used as the second feature data P2 of each target welding robot; the operation time required for each target welding robot to complete one corresponding welding operation is obtained and the average operation time is calculated, and the average operation time is used as the third feature data P3 of each target welding robot. In an embodiment of the present invention, the system first obtains a plan view of the work area of the automated welding production line and marks the target welding robots identified in step S100 on the plan view. Then, the system extracts the feature data of each target welding robot, including: collision warning threshold (first feature data P1), average trajectory adjustment response time (second feature data P2), and average single operation time (third feature data P3). These data provide input for subsequent trajectory association judgment and dynamic obstacle avoidance trajectory generation. The process of determining the correlation between the work trajectories of every two target welding robots in step S300 includes: Step S301: Query the work area information of the target welding robot by dividing the production line workstation data. If the two target welding robots are not in the same work area, it is determined that there is no work trajectory association between the two target welding robots. If the two target welding robots are in the same work area and complete different welding processes in the same work area, feature data is extracted for the two target welding robots. Step S302: Based on the welding sequence of the two target welding robots in the same work area, the two target welding robots are designated as target welding robot X1 and target welding robot X2; the welding process of target welding robot X1 is located before the welding process of target welding robot X2; the first feature data P(X1,1), the second feature data P(X1,2), and the third feature data P(X1,3) of target welding robot X1, and the first feature data P(X2,1), the second feature data P(X2,2), and the third feature data P(X2,3) of target welding robot X2 are obtained respectively; the average trajectory adjustment speed V of the welding robot is calculated by collecting the motion data of the welding robot through sensors, and the shortest working distance S between target welding robot X1 and target welding robot X2 is calculated by using the coordinate data of the work area layout plan; Step S303: In the historical operation monitoring data of the automated welding production line, when the cumulative value of the working condition of the target welding robot X1 reaches the first feature data P(X1,1), calculate the average cumulative value of the working condition of the corresponding target welding robot X2 P(X2,u). If P(X2,u)≥P(X2,1), it is determined that the target welding robot X1 and the target welding robot X2 satisfy the operation trajectory association, and the target welding robot X2 is the trajectory association robot of the target welding robot X1. If P(X2,u) < P(X2,1), using the second feature data P(X1,2), the shortest working distance S, and the average trajectory adjustment speed V, calculate the shortest time for the target welding robot X1 to move to a non-crossing risk working position after completing one trajectory adjustment. If, after completing one work trajectory adjustment, the target welding robot X1 moves to a work position with no cross-risk within the shortest time T, predict the cumulative work condition value P(X2,r) corresponding to the target welding robot X2, where, ,in, Let P(X2,r) represent the number of welding operations that the target welding robot X2 can complete within the shortest time T. If P(X2,r)≥P(X2,1), then it is determined that the target welding robot X1 and the target welding robot X2 satisfy the operation trajectory association, and the target welding robot X2 is the trajectory association robot of the target welding robot X1. In an embodiment of the present invention, the system performs a work trajectory association judgment on every two target welding robots based on the feature data of the target welding robots. The judgment process includes: querying the production line workstation division data to determine whether the two welding robots are in the same work area; if they are in the same area and there is a process sequence relationship, then their feature data is extracted; by calculating the shortest working distance S and the average trajectory adjustment speed V, and combining the cumulative working condition value of X2 when X1 reaches the warning threshold in historical data, it is determined whether the two meet the trajectory association conditions; if they meet the conditions, then X2 is determined to be the trajectory associated robot of X1. Step S400 includes: Step S401: Obtain the trajectory-associated robots for each target welding robot; in the work area layout diagram, draw the shortest movement path between each target welding robot and its corresponding trajectory-associated robots, and mark the other work stations passed through by the shortest movement path; if the number of other work stations is less than the work station number threshold, then take each target welding robot as the trajectory starting point, generate a dynamic obstacle avoidance operation trajectory in the work area layout diagram that covers the corresponding trajectory-associated robots and other work stations, and store the dynamic obstacle avoidance operation trajectory; Step S402: If the number of other workstations is greater than the workstation threshold, then take each target welding robot as the trajectory starting point, and in the work area layout diagram, combine the shortest working movement distance S and the average trajectory adjustment speed V of the welding robot to generate the shortest dynamic obstacle avoidance operation trajectory from each target welding robot to each trajectory-related robot, and store the shortest dynamic obstacle avoidance operation trajectory. In an embodiment of the present invention, a dynamic obstacle avoidance trajectory is generated in the work area layout diagram based on the distribution of the trajectory-associated robots of each target welding robot. Specifically, the system first calculates the shortest movement path between each target welding robot and its trajectory-associated robot, and counts the number of other workstations passed through by the shortest movement path. If the number of workstations is less than a preset threshold, a dynamic obstacle avoidance trajectory covering multiple workstations is generated. If the number of workstations is greater than the threshold, the shortest obstacle avoidance trajectory between two points is generated. All generated trajectories are stored in the system for subsequent scheduling. In step S500, when accumulating the real-time welding operation status of each target welding robot, the accumulated value of the real-time welding operation status of each target welding robot is cleared to zero after the target welding robot completes the obstacle avoidance cooperative operation according to the dynamic obstacle avoidance operation trajectory. In an embodiment of the present invention, the real-time welding operation status of each target welding robot is accumulated; when the accumulated value of the operation status of a target welding robot reaches its preset collision warning threshold, the system automatically generates a dynamic coordination command for the robot and extracts a dynamic obstacle avoidance operation trajectory starting from the robot from the database; after the obstacle avoidance operation is completed, the system clears the accumulated value of the robot's operation status to zero and prepares for the next round of monitoring and scheduling. The system transmits the generated dynamic coordination instructions and corresponding dynamic obstacle avoidance operation trajectories to the target welding robot's operating terminal via industrial communication protocols (such as OPC UA, Modbus TCP, etc.); the robot adjusts its operation trajectory according to the instructions to complete the safe and collaborative welding operation; the system monitors the execution process in real time and updates the status data after the operation is completed, forming a closed-loop supervision process; In order to better implement the above method, a multi-sensor-based intelligent monitoring system for welding robots is also provided. The intelligent monitoring system includes: a risk identification module, a feature extraction module, an associated trajectory judgment module, a dynamic obstacle avoidance trajectory generation module, an instruction generation module, and a transmission control module. The risk identification module is used to acquire historical operation monitoring data of the automated welding production line, capture the operation information of each welding robot, integrate and process the data collected by the sensors, and identify target welding robots with trajectory intersection risks based on the collected data. The feature extraction module is used to obtain a planar layout of the work area, annotate the target welding robot, and extract the feature data of the target welding robot. The associated trajectory judgment module is used to perform correlation analysis on the operation trajectories between every two target welding robots based on the feature data of the target welding robots, and to determine the associated robots of each target welding robot's trajectory. The dynamic obstacle avoidance trajectory generation module is used to generate and store several segments of dynamic obstacle avoidance operation trajectories in the work area layout map based on the distribution of trajectory-related robots. The instruction generation module is used to accumulate the welding operation conditions of the target welding robot. When the accumulated value reaches the collision warning threshold, it generates dynamic coordination instructions and extracts the corresponding dynamic obstacle avoidance operation trajectory. The transmission control module is used to transmit dynamic coordination commands and dynamic obstacle avoidance operation trajectories to the operation terminal of the corresponding target welding robot, assisting it in completing safe and collaborative welding operations. The risk identification module includes: a sensor data acquisition unit and a risk target identification unit; The sensor data acquisition unit is used to collect the equipment parameters and operating data of each welding robot in the automated welding production line in real time through sensors, lock the welding operation trajectory of each welding robot and capture its working space range during the operation. The risk target identification unit is used to determine whether there is an overlapping area in the workspace of each welding robot. If there is, the robot is marked as the target welding robot, and the probability of its trajectory intersecting with other robots during operation is calculated. The correlation trajectory judgment module includes: a work area query unit and a correlation analysis unit; The work area query unit is used to query the work area information of the target welding robot based on the production line workstation division data, and to determine whether two target welding robots are in the same work area. The correlation analysis unit is used to extract feature data of target welding robots in the same work area, and calculate and determine whether there is a correlation between the two welding robots by combining the welding process sequence, the shortest working distance, and the average trajectory adjustment speed, and then identify the trajectory-correlated robot.
[0020] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for intelligent monitoring of welding robots based on multiple sensors, characterized in that: The intelligent supervision method includes: Step S100: Obtain historical operation monitoring data of the automated welding production line and capture the operation information of each welding robot in the production line; transmit the operation information to the welding supervision terminal in a unified manner; the welding supervision terminal integrates and processes the welding robot operation information of the production line through sensors and captures target welding robots with the risk of trajectory intersection. Step S200: Obtain the work area layout plan of the automated welding production line, mark the target welding robot in the work area layout plan, and extract feature data of the target welding robot; Step S300: Based on the feature data of each target welding robot, perform correlation judgment on the operation trajectory between every two target welding robots to determine the trajectory associated robot of each target welding robot; Step S400: Based on the distribution of the target welding robots associated with their trajectories, generate several segments of dynamic obstacle avoidance trajectories in the work area layout diagram, and store the dynamic obstacle avoidance trajectories. Step S500: Accumulate the welding operation conditions corresponding to each target welding robot; when the accumulated value of the operation condition corresponding to a target welding robot reaches the collision warning threshold of the target welding robot, generate the dynamic coordination command of the target welding robot; at the same time, extract the dynamic obstacle avoidance operation trajectory with the target welding robot as the trajectory starting point. Step S600: Simultaneously transmit the dynamic coordination command and the dynamic obstacle avoidance operation trajectory to the operation terminal of the corresponding target welding robot, and control and assist each target welding robot to complete the safe and collaborative welding operation.
2. The intelligent monitoring method for welding robots based on multiple sensors according to claim 1, characterized in that: Step S100 includes: Step S101: Capture the equipment parameters and operating data of each welding robot in the automated welding production line through sensors, lock the welding operation trajectory of each welding robot according to the equipment parameters of each welding robot, and capture the working space range of each welding robot in the process of executing the corresponding welding operation trajectory. Step S102: If the workspace of a certain welding robot overlaps with the workspace of other welding robots, the welding robot is designated as the target welding robot; wherein, the overlapping area is the work area where trajectory intersection is likely to occur when multiple robots work together; the probability of trajectory intersection between each target welding robot and other robots during the execution of the corresponding welding operation trajectory is obtained respectively.
3. The intelligent monitoring method for welding robots based on multiple sensors according to claim 2, characterized in that: Based on the trajectory intersection probability of each target welding robot, a collision warning threshold is set for each target welding robot to trigger the generation of dynamic coordination commands.
4. The intelligent monitoring method for welding robots based on multiple sensors according to claim 1, characterized in that: Step S200 includes: The collision warning threshold of each target welding robot when the dynamic coordination command is triggered is obtained, and the collision warning threshold is used as the first feature data P1 of each target welding robot; the response time of each target welding robot in each completion of the work trajectory adjustment is obtained and the average response time is calculated, and the average response time is used as the second feature data P2 of each target welding robot; the operation time required for each target welding robot to complete one corresponding welding operation is obtained and the average operation time is calculated, and the average operation time is used as the third feature data P3 of each target welding robot.
5. The intelligent monitoring method for welding robots based on multiple sensors according to claim 1, characterized in that: Step S300, which involves determining the correlation between the work trajectories of every two target welding robots, includes: Step S301: Query the work area information of the target welding robot by dividing the production line workstation data. If the two target welding robots are not in the same work area, it is determined that there is no work trajectory association between the two target welding robots. If the two target welding robots are in the same work area and complete different welding processes in the same work area, feature data is extracted for the two target welding robots. Step S302: Based on the welding sequence of the two target welding robots in the same work area, the two target welding robots are designated as target welding robot X1 and target welding robot X2; the welding process of target welding robot X1 is located before the welding process of target welding robot X2; the first feature data P(X1,1), the second feature data P(X1,2), and the third feature data P(X1,3) of target welding robot X1, and the first feature data P(X2,1), the second feature data P(X2,2), and the third feature data P(X2,3) of target welding robot X2 are obtained respectively; the average trajectory adjustment speed V of the welding robot is calculated by collecting the motion data of the welding robot through sensors, and the shortest working distance S between target welding robot X1 and target welding robot X2 is calculated by using the coordinate data of the work area layout plan; Step S303: In the historical operation monitoring data of the automated welding production line, when the cumulative value of the working condition of the target welding robot X1 reaches the first feature data P(X1,1), calculate the average cumulative value of the working condition of the corresponding target welding robot X2 P(X2,u). If P(X2,u)≥P(X2,1), it is determined that the target welding robot X1 and the target welding robot X2 satisfy the operation trajectory association, and the target welding robot X2 is the trajectory association robot of the target welding robot X1. If P(X2,u) < P(X2,1), using the second feature data P(X1,2), the shortest working distance S, and the average trajectory adjustment speed V, calculate the shortest time for the target welding robot X1 to move to a non-crossing risk working position after completing one trajectory adjustment. If, after completing one work trajectory adjustment, the target welding robot X1 moves to a work position with no cross-risk within the shortest time T, predict the cumulative work condition value P(X2,r) corresponding to the target welding robot X2, where, ,in, Let P(X2,r) represent the number of welding operations that target welding robot X2 can complete within the shortest time T. If P(X2,r)≥P(X2,1), then it is determined that target welding robot X1 and target welding robot X2 satisfy the operation trajectory association, and target welding robot X2 is the trajectory-associated robot of target welding robot X1.
6. The intelligent monitoring method for welding robots based on multiple sensors according to claim 1, characterized in that: Step S400 includes: Step S401: Obtain the trajectory-associated robots for each target welding robot; in the work area layout diagram, draw the shortest movement path between each target welding robot and its corresponding trajectory-associated robots, and mark the other work stations passed through by the shortest movement path; if the number of other work stations is less than the work station number threshold, then take each target welding robot as the trajectory starting point, generate a dynamic obstacle avoidance operation trajectory in the work area layout diagram that covers the corresponding trajectory-associated robots and other work stations, and store the dynamic obstacle avoidance operation trajectory; Step S402: If the number of other workstations is greater than the workstation threshold, then take each target welding robot as the trajectory starting point, and in the work area layout diagram, combine the shortest working movement distance S and the average trajectory adjustment speed V of the welding robot to generate the shortest dynamic obstacle avoidance operation trajectory from each target welding robot to each trajectory-related robot, and store the shortest dynamic obstacle avoidance operation trajectory.
7. The intelligent monitoring method for welding robots based on multiple sensors according to claim 1, characterized in that: In step S500, when accumulating the real-time welding operation status of each target welding robot, the accumulated value of the real-time welding operation status of each target welding robot is cleared to zero after the target welding robot completes the obstacle avoidance cooperative operation according to the dynamic obstacle avoidance operation trajectory.
8. A multi-sensor-based intelligent monitoring system for welding robots, used to execute the multi-sensor-based intelligent monitoring method for welding robots according to any one of claims 1-7, characterized in that: The intelligent monitoring system includes: a risk identification module, a feature extraction module, a correlation trajectory judgment module, a dynamic obstacle avoidance trajectory generation module, an instruction generation module, and a transmission control module; The risk identification module is used to acquire historical operation monitoring data of the automated welding production line, capture the operation information of each welding robot, integrate and process the data collected by the sensors, and identify target welding robots with trajectory intersection risks based on the collected data. The feature extraction module is used to obtain a planar layout of the work area, mark the target welding robot, and extract the feature data of the target welding robot. The associated trajectory judgment module is used to perform correlation analysis on the operation trajectory between every two target welding robots based on the feature data of the target welding robots, and to determine the trajectory associated robot of each target welding robot. The dynamic obstacle avoidance trajectory generation module is used to generate and store several segments of dynamic obstacle avoidance operation trajectory in the work area plan layout map according to the distribution of trajectory-associated robots. The instruction generation module is used to accumulate the welding operation conditions of the target welding robot. When the accumulated value reaches the collision warning threshold, it generates a dynamic coordination instruction and extracts the corresponding dynamic obstacle avoidance operation trajectory. The transmission control module is used to transmit dynamic coordination commands and dynamic obstacle avoidance operation trajectories to the operation terminal of the corresponding target welding robot, assisting it in completing safe and collaborative welding operations.
9. The intelligent monitoring method for welding robots based on multiple sensors according to claim 8, characterized in that: The risk identification module includes: a sensor data acquisition unit and a risk target identification unit; The sensor data acquisition unit is used to collect the equipment parameters and operating data of each welding robot in the automated welding production line in real time through sensors, lock the welding operation trajectory of each welding robot and capture its working space range during the operation. The risk target identification unit is used to determine whether there is an overlapping area in the workspace of each welding robot. If there is, the robot is marked as the target welding robot, and the probability of its trajectory intersecting with other robots during operation is calculated.
10. The intelligent monitoring method for welding robots based on multiple sensors according to claim 8, characterized in that: The associated trajectory determination module includes: a work area query unit and an association analysis unit; The work area query unit is used to query the work area information of the target welding robot based on the production line workstation division data, and to determine whether two target welding robots are in the same work area. The correlation analysis unit is used to extract feature data from target welding robots in the same work area, and calculate and determine whether there is a correlation between the two welding robots by combining the welding process sequence, the shortest working distance, and the average trajectory adjustment speed, and thus identify the trajectory-correlated robot.