Multi-robot cooperative control method based on new energy street lamp production line data

Abstract

The application relates to the technical field of electric digital data processing, and particularly discloses a multi-robot cooperative control method based on new energy street lamp production line data, which comprises the following steps: a central control port receives a new energy street lamp assembly total task, performs task decomposition, obtains various to-be-assembled function domain tasks, and sends a preliminary calling signal to each assembly robot in the production line. After the robot receives the signal, the real-time execution data of the robot is acquired to determine the real-time load evaluation value, and the load state data is fed back to the central control port. The central control port cooperatively adjusts the to-be-assembled function domain tasks of the assembly robot according to the data. Meanwhile, the task execution and execution state of the robot are monitored in real time, so that the production process can be smoothly carried out. This process realizes accurate control of the assembly robot and efficient allocation of the task, and improves the automation degree and production efficiency of the new energy street lamp production line.

Description

Technical Field

[0001] This invention relates to the field of robot collaborative control technology, specifically a multi-robot collaborative control method based on data from a new energy street light production line. Background Technology

[0002] Currently, multi-robot control in street light production lines primarily relies on pre-programmed, fixed logic to coordinate robot movements. Task allocation is based on the initial planning of production tasks, and robots execute tasks according to a preset sequence and trajectory, with little dynamic adjustment or only fine-tuning based on simple sensor feedback. In the production of complex new energy street lights, this traditional method suffers from problems such as unreasonable task allocation, insufficient real-time adjustment, and inefficient resource utilization, making it difficult to adapt to the rapid switching between different street light models, resulting in reduced production efficiency and product quality. Therefore, there is an urgent need to develop a highly automated and flexible new energy street light production line capable of multi-robot collaborative operation to improve production efficiency and product quality.

[0003] For example, the invention patent with announcement number CN118394009B discloses a production assembly control method for medium-intensity obstruction lights, which relates to the field of automation control. The method includes: obtaining configuration information of Q nodes; extracting the target assembly quantity; querying historical assembly control schemes to obtain multiple historical assembly control schemes; performing fitness calculation, using the historical assembly control scheme corresponding to the stage with the maximum fitness value as the adjustment target, adjusting the remaining multiple historical assembly control schemes, constraining the adjustment process with the transfer capacity of Q nodes and the adjustable space of assembly efficiency of Q nodes, and obtaining multiple updated historical assembly control schemes; further adjusting and optimizing to generate the target assembly control scheme; and controlling the Q assembly nodes.

[0004] For example, invention patent CN116339260A discloses a production time adjustment and analysis method and system for flexible manufacturing in a "lights-out" factory, relating to the field of process control. It includes: acquiring the number of machines and their cycle times on a flexible production line; acquiring preset product production indicators, each process flow, and the production time of each process flow; obtaining the product's cycle time based on the production indicators and machine cycle times; improving each process flow; and observing and recording the production time of the improved process flows.

[0005] Based on the above technical solutions, it was found that most existing production line control technologies rely on static process breakdown based on existing historical data and single process constraints to control the production line. Due to the complexity and diversity of processes in street light production lines, multiple robots need to cooperate on the production line. However, existing technologies lack flexible control over real-time data of the production line, which easily leads to problems such as mismatched task allocation and low multi-robot collaboration rate, greatly reducing the efficiency and quality of the production line. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a multi-robot collaborative control method based on data from a new energy street light production line, which can effectively solve the problems mentioned in the background technology.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a multi-robot collaborative control method based on data from a new energy street light production line, comprising: a central control port receiving the total assembly task of the street lights, decomposing the task to obtain the tasks of each functional domain to be assembled, and generating a pre-call signal to the signal receiving port of each assembly robot in the production line; after receiving the pre-call signal, the signal receiving port acquiring the real-time execution data of each assembly robot, determining the real-time load assessment value of each assembly robot, feeding back the real-time load status data of each assembly robot to the central control port, and determining whether to perform collaborative adjustment of the assembly robots; based on the real-time load status data, coordinating the tasks of each assembly robot to be assembled in the functional domain through the central control port; and monitoring the current task execution status of each assembly robot in real time and feeding back the execution status of each assembly robot.

[0008] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0009] (1) This invention provides a multi-robot collaborative control method based on data from a new energy street light production line. The central control port receives the total assembly task for the new energy street lights, decomposes the task into tasks for each functional domain to be assembled, and sends a pre-call signal to each assembly robot in the production line. Upon receiving the signal, the robot acquires its own real-time execution data to determine the real-time load assessment value and feeds back the load status data to the central control port. The central control port then uses this data to coordinate the assembly robots' tasks for each functional domain to be assembled. Simultaneously, it monitors the robots' task execution and execution status in real time to ensure the smooth progress of the production process. This process achieves precise control of the assembly robots and efficient task allocation, improving the automation level and production efficiency of the new energy street light production line.

[0010] (2) This invention obtains the real-time execution data of each assembly robot and determines the real-time load assessment value of each assembly robot. This is used to monitor the robot's load status in real time. The central control system can dynamically adjust the task allocation, transferring some tasks from high-load robots to idle robots, thereby achieving a balanced allocation of tasks and improving overall production efficiency. This avoids low production efficiency caused by unreasonable task allocation or excessive robot load, ensuring the smooth operation of the production line. It enables the central control port to quickly respond to changes and demands in the production process, such as a sudden increase in the workload or robot malfunctions, and adjusts the task allocation and execution strategy in a timely manner to ensure the continuity and stability of production.

[0011] (3) This invention obtains the task execution data of the assembly robot, determines the execution quality indicators of the assembly robot, monitors the execution quality of the assembly robot in real time, ensures that each assembly step meets the predetermined quality standards, promptly identifies and corrects quality problems, and improves the product qualification rate. It reduces production interruptions and rework caused by quality problems, ensures the smooth progress of the production process, and improves overall production efficiency. Based on the feedback of the execution quality indicators, it adjusts and optimizes the robot's operating parameters and processes, improving the stability and reliability of the production process. Attached Figure Description

[0012] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0013] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0014] Figure 2 This is a system architecture diagram for a fully automated flexible production line for new energy streetlights.

[0015] Figure 3 This is a flowchart of a multi-robot collaborative operation.

[0016] Figure 4 This is a diagram showing the functional modules of the central control system.

[0017] Figure 5 This is a schematic diagram of a multi-robot collaborative control process.

[0018] Figure 6 Flowchart for assigning tasks to functional domains to be assembled.

[0019] The attached figures are labeled 1 and R1 to R4, which represent the assembly robot numbers, respectively. Detailed Implementation

[0020] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0021] Reference Figure 1As shown, the present invention provides a multi-robot collaborative control method based on data from a new energy street light production line, including: a central control port receiving the total assembly task of the street lights in the production line, decomposing the task to obtain the functional domain tasks to be assembled, and generating a pre-call signal to the signal receiving port of each assembly robot in the production line.

[0022] It should be noted that the street light production line includes material conveying modules, processing modules, assembly modules, testing modules, and packaging modules, as well as a central control system. For example... Figure 2 As shown, Figure 2 This is a system architecture diagram of a fully automated flexible production line for new energy streetlights. The assembly module is responsible for assembling the various components into a complete new energy streetlight, including: a lamp assembly station (integrating LED lamps and driver circuits); an electronic control system assembly station (installing control circuits, batteries, etc.); a solar panel installation station (fixing solar panels and connecting circuits); and a complete unit assembly station (integrating all components onto a bracket). When different models of streetlights need to be produced, the central control system regenerates the production plan, automatically adjusts the process parameters of each module and the robot's working program, achieving rapid switching without manual adjustment.

[0023] In this embodiment of the invention, the robots coordinate and adjust, and the specific process is as follows: Figure 3 As shown, Figure 3 The flowchart for multi-robot collaborative operation is as follows. The multi-robot collaborative operation method mainly includes five steps: task decomposition, resource allocation, path planning, collaborative execution, and exception handling.

[0024] Specifically, task decomposition involves breaking down the complex task of producing new energy streetlights into multiple sub-tasks: task characteristic-based decomposition, which categorizes tasks according to their functional characteristics, such as material handling, component installation, and precision operation; time-based decomposition, which arranges tasks according to their execution order to ensure task dependencies; and spatial decomposition, which allocates tasks according to their spatial location to reduce interference between robots.

[0025] The collaborative execution specifically includes: a real-time communication mechanism (establishing a real-time communication network between robots to share status information); a synchronous control strategy (adopting a synchronous control strategy to ensure that each robot works in coordination according to the predetermined plan); a feedback adjustment mechanism (adjusting the execution strategy in real time based on feedback information during the execution process); and precise positioning technology (using a high-precision visual positioning system to ensure the accuracy of collaboration).

[0026] It should be noted that, in this embodiment of the invention, a central control port also generates optimal motion paths for each robot. Specifically: single-robot path planning: generating collision-free motion paths for each robot; multi-robot collaborative path planning: considering the mutual influence between robots to avoid conflicts; spatiotemporal coordination algorithm: based on the spatiotemporal coordination algorithm, achieving efficient collaboration among multiple robots in the same workspace; path optimization: optimizing paths to reduce execution time and energy consumption.

[0027] In this embodiment of the invention, the central control port is the core of the entire production line, specifically as follows: Figure 4 As shown, Figure 4 This is a diagram showing the functional modules of the central control system, used to achieve production scheduling: generating production plans based on order requirements and allocating tasks to each workstation; quality monitoring: monitoring the production process in real time and promptly identifying and handling quality issues; data management: collecting and analyzing production data to provide a basis for production optimization; and robot control: coordinating the collaborative operation of multiple robots.

[0028] After receiving the pre-call signal, the signal receiving port obtains the real-time execution data of each assembly robot, determines the real-time load assessment value of each assembly robot, feeds back the real-time load status data of each assembly robot to the central control port, and determines whether to perform coordinated adjustment of the assembly robots.

[0029] The coordinated adjustment process of the assembly robots described above is as follows: Figure 5 As shown, Figure 5 This diagram illustrates a multi-robot collaborative control process. The real-time load assessment value of the assembly robot is monitored and compared with a preset threshold. Units with a load less than or equal to the threshold are considered idle and require no collaboration; units with a load greater than the threshold are considered high-load and require collaborative adjustment. During collaborative adjustment, the load deviation is compared with the preset threshold. If the deviation is less than or equal to the threshold, the robot is overloaded, triggering the next round of allocation; otherwise, it is overloaded, initiating master-slave collaborative adjustment. The high-load unit becomes the master robot, gaining task scheduling rights. It selects the idle unit with the smallest load assessment value in the same attribute work domain as the slave robot and broadcasts a master-slave change command. The master robot splits tasks, migrating subtasks with waiting times exceeding a defined time limit to the slave robots. If a slave robot's task exceeds the permitted limit, a second slave robot is selected. The master robot monitors the slave robot's status. If its real-time motor speed is lower than the reference speed for an accumulated time exceeding a defined time limit, it issues a compensation command.

[0030] Based on real-time load status data, the central control port coordinates and adjusts the assembly robots' functional domain tasks.

[0031] The coordinated adjustment of the aforementioned functional domain tasks to be assembled is specifically as follows: Figure 6 As shown, Figure 6A flowchart for assigning tasks to functional domains to be assembled is provided. The central control port obtains the current load status of the assembly robot based on the task dependency matrix. During high load, a master-slave task queue is created, the auxiliary robot takes over the slave tasks, and the master robot increases the production cycle. When idle, tasks are executed. After completion, feedback data is used to determine the execution quality indicators. If the indicators are lower than the reference indicators, the robot's self-calibration program is triggered. The joint zero points are recalibrated using laser positioning landmarks, and the calibration program is configured according to the quality deviation mapping laser sensor to adapt the scanning frequency. The execution status is then fed back.

[0032] The system monitors the current task execution status of each assembly robot in real time and provides feedback on the execution status of each assembly robot.

[0033] Specifically, the task is broken down, and the detailed analysis process is as follows:

[0034] The task decomposition includes primary decomposition and secondary decomposition.

[0035] The first-level decomposition specifically involves dividing the overall street light assembly task into several production functional domains based on functional characteristics, including the lamp assembly functional domain, the electronic control system assembly functional domain, the solar module installation functional domain, and the complete machine assembly functional domain. Each functional domain corresponds to an independent robot operation unit.

[0036] The above functional division aims to achieve specialized division of labor in assembly work, improve assembly efficiency and quality, and facilitate management and control. The lighting fixture assembly functional domain includes sub-queues for LED chip mounting, optical lens integration, and lighting fixture housing encapsulation. The electrical control system assembly functional domain includes sub-queues for battery rack mounting and main control board installation. The solar module installation functional domain includes sub-queues for photovoltaic panel bracket mounting and solar panel hoisting. The complete unit assembly functional domain includes sub-queues for pole docking and integration of various functional domains.

[0037] The second-level decomposition specifically involves: arranging the tasks of each functional domain to be assembled in the order of execution, ensuring that each task satisfies a preset task dependency matrix, generating a queue of sub-tasks with time-series labels, and allocating each production functional domain according to its spatial location, generating a queue of sub-tasks with spatial labels.

[0038] The task dependency matrix is ​​a tool used to clearly describe the sequence and interdependencies between assembly tasks. In the production line of new energy streetlights, it ensures a rational and orderly assembly process. It takes the form of a table or matrix, where rows and columns represent different tasks, and matrix elements represent the dependency between task i and task j, with values ​​of 1 or 0. For example, the intersection of the row corresponding to the bracket fixing task and the column corresponding to the solar panel installation task has a value of 1, indicating that the bracket fixing must be completed before the solar panel installation can proceed. Other similar dependencies, such as the need to complete the wiring of the electrical control system before assembling the lighting modules, are also clearly marked in the matrix. The matrix's function is to optimize task scheduling, ensure the rational allocation and efficient utilization of resources, and guarantee that the assembly process meets technological requirements, preventing assembly errors or rework due to incorrect task sequence.

[0039] Furthermore, the real-time load assessment value of each assembly robot is determined. The specific determination process is as follows:

[0040] The real-time execution data of each assembly robot includes the real-time motor operating load, the real-time production cycle time lag deviation, the real-time task queue backlog, and the real-time end effector occupancy rate. This real-time execution data can be extracted from the execution logs of each assembly robot.

[0041] The real-time motor operating load, real-time production cycle time lag deviation, real-time task queue backlog, and real-time end effector occupancy rate of each assembly robot are normalized. The normalization results are then weighted and aggregated to obtain the real-time load evaluation value of each assembly robot. The specific analysis process is as follows:

[0042]

[0043] In the formula, LF f Let f be the real-time load assessment value of the f-th assembly robot, where f is the number of each assembly robot, f=[1,F], F is the total number of assembly robots, and t is the time variable. t0 is the start time of the monitoring cycle, t1 is the end time of the monitoring cycle, and ML ft For the real-time motor operating load of the f-th assembly robot at time t, PLB ft SQ represents the real-time production cycle time lag deviation of the f-th assembly robot at time t. ft Let OP be the real-time task queue backlog of the f-th assembly robot at time t. ftLet b1 be the real-time end effector occupancy rate of the f-th assembly robot at time t, b2 be the weight element corresponding to the real-time motor operating load predefined in the robot management information database, b3 be the weight element corresponding to the real-time production cycle lag deviation predefined in the robot management information database, b4 be the weight element corresponding to the real-time task queue backlog predefined in the robot management information database, and b5 be the weight element corresponding to the real-time end effector occupancy rate predefined in the robot management information database.

[0044] It should be explained that the above normalization process specifically involves first normalizing the numerical range of motor operating load, production cycle lag deviation, task queue backlog, and end effector occupancy rate (mapping the values ​​of each physical quantity to a unified range to eliminate differences in magnitude), and then normalizing the dimensions (removing units through ratio calculations of their own ranges to achieve dimensionlessness). This transforms physical quantities with inconsistent units and numerical scales into dimensionless forms with consistent numerical ranges.

[0045] It should be explained that the aforementioned production cycle time lag deviation refers to the difference between the actual cycle time of the assembly robot completing the task and the predetermined production cycle time. It is expressed as the difference between the predetermined and actual production cycle times when the actual cycle time is shorter than the predetermined cycle time. For example, if the predetermined cycle time is one specific assembly action per minute, but the actual completion time is longer than the predetermined time, this difference is the real-time production cycle time lag deviation. The end effector occupancy rate refers to the usage status of the assembly robot's end effectors (such as fixtures, welding torches, etc.) at the current moment, that is, the proportion of time the end effector is in operation out of the total time. It reflects the busyness level of the end effector.

[0046] The weight elements corresponding to real-time motor operating load, real-time production cycle time lag deviation, real-time task queue backlog, and real-time end effector occupancy rate are all extracted from the robot management information database. The mapping relationship can be one-to-one or many-to-one. For example, the real-time motor operating load, real-time production cycle time lag deviation, real-time task queue backlog, and real-time end effector occupancy rate are respectively mapped to the weight elements corresponding to real-time motor operating load, real-time production cycle time lag deviation, real-time task queue backlog, and real-time end effector occupancy rate preset in the robot management information database to form a mapping set. Substituting the real-time motor operating load, real-time production cycle time lag deviation, real-time task queue backlog, and real-time end effector occupancy rate into the mapping set yields the weight elements corresponding to real-time motor operating load, real-time production cycle time lag deviation, real-time task queue backlog, and real-time end effector occupancy rate.

[0047] In this embodiment, multivariate analysis of real-time motor operating load, real-time production cycle time lag deviation, real-time task queue backlog, and real-time end effector occupancy rate is used to consider the correlation between these parameters. An increase in real-time motor operating load may lead to a decrease in robot operating speed, thereby increasing the real-time production cycle time lag deviation. At the same time, it will also relatively increase the end effector occupancy rate. Furthermore, due to the slower task execution, the real-time task queue backlog may gradually increase, thus greatly increasing the real-time load assessment value of the assembly robot. If the real-time production cycle time lag deviation is large and not adjusted for a long time, the real-time task queue backlog will continue to increase. This may also lead to the motor running at high load for a long time, and the end effector occupancy rate will remain high, similarly increasing the real-time load assessment value of the assembly robot.

[0048] Specifically, the process for determining whether to perform coordinated adjustments to the assembly robot is as follows:

[0049] The real-time load assessment value of each assembly robot is compared with the predefined load assessment allowable threshold. If the real-time load assessment value of a certain assembly robot is less than or equal to the load rate assessment allowable threshold, no coordinated adjustment of the assembly robot is required. The assembly robot is marked as an idle unit, and the execution status data of the idle unit is fed back to the central control port.

[0050] If the real-time load assessment value of an assembly robot is greater than the load assessment allowable threshold, then the assembly robot needs to be coordinated and adjusted. The assembly robot is marked as a high-load unit, and the execution status data of the high-load unit is fed back to the central control port. At the same time, the difference between the real-time load assessment value of the assembly robot and the load assessment allowable threshold is processed to obtain the load deviation of the assembly robot, and the overflow load state of the assembly robot is determined, and coordinated adjustment is performed.

[0051] The aforementioned execution status data includes, but is not limited to, basic feedback: the spatial coordinates of the assembly robot, the moving speed of the assembly robot, and the task progress of the assembly robot, as well as key feedback: the torque of the assembly robot, the temperature of the assembly robot, etc.

[0052] By continuously monitoring the real-time load assessment values ​​of assembly robots and comparing them with predefined load rate assessment thresholds, the load status of each robot can be grasped in a timely and accurate manner. This avoids low production efficiency or quality problems caused by unreasonable task allocation or excessive robot load. When a robot is overloaded, timely feedback is provided and a collaborative adjustment mechanism is activated to dynamically optimize task allocation. Some tasks are transferred to idle robots, making task allocation more balanced and reasonable, improving overall production efficiency, and preventing the operation of the entire production line from being affected by the overload of some robots.

[0053] Furthermore, the overflow load state of the assembly robot is determined, and coordinated adjustments are made. The specific determination process is as follows:

[0054] The load deviation of the assembly robot is compared with a predefined load deviation threshold. If the load deviation of the assembly robot is less than or equal to the load deviation threshold, the overflow load state of the assembly robot is determined to be overloaded, and the next round of collaborative allocation is triggered first to maintain the current task backlog queue of the assembly robot.

[0055] If the load deviation of the assembly robot is greater than the load deviation threshold, the overflow load state of the assembly robot is determined to be overloaded, and master-slave collaborative adjustment is initiated. The high-load unit is automatically upgraded to master robot and obtains task scheduling rights. At the same time, the real-time load evaluation value of each idle unit in the same attribute work domain is extracted, and the idle unit with the smallest real-time load evaluation value is selected as the auxiliary robot. The central control port broadcasts the master-slave relationship change instruction to the auxiliary robot.

[0056] It should be explained that the aforementioned "same attribute work domain" refers to similar areas within a production line, divided according to the nature, function, or resource requirements of the tasks. Tasks within these areas share similar attributes, such as requiring specific types of robots, similar tools or resources, similar operational steps, or the same environmental requirements. For example, within the same attribute work domain, all tasks involve similar assembly operations, such as all electrical connection tasks or all mechanical assembly tasks. Robots or equipment within this area have the same or similar functions, such as high-precision positioning and welding capabilities.

[0057] Based on task scheduling authority, the master robot assigns tasks to the auxiliary robots. Task scheduling authority, in multi-robot collaborative operations, refers to the decision-making power of a particular robot or control system component regarding task allocation and execution order. The entity possessing task scheduling authority is responsible for rationally allocating tasks to other robots and determining the task execution order based on factors such as current production needs, robot load status, and task priority.

[0058] In the master-slave collaborative adjustment mechanism, the central control port grants the master robot the task scheduling authority. Based on its own task backlog and load status, the master robot decides which tasks need to be migrated to the auxiliary robot and assigns specific tasks to the auxiliary robot for execution.

[0059] By further refining the load management mechanism, it not only determines whether the assembly robot is overloaded, but also more accurately determines whether the robot is in a heavy or overloaded state by comparing the load deviation with a predefined load deviation threshold, providing a more accurate basis for subsequent coordinated adjustment measures. When the robot is overloaded, the high-load unit automatically becomes the master robot and gains task scheduling rights. At the same time, a suitable idle unit is selected as the auxiliary robot based on the real-time load assessment value. This clear division of master and slave relationships makes task allocation more orderly, collaborative operations more targeted, and improves the efficiency and coordination of multi-robot collaborative operations.

[0060] Specifically, the main robot assigns tasks to the auxiliary robot, and the detailed analysis process is as follows:

[0061] The main robot splits its own task backlog into sub-backlog tasks and obtains the waiting time of each sub-backlog task. The waiting time can be extracted from the main robot's execution record. Sub-backlog tasks with a waiting time longer than the preset waiting time limit are recorded as migration tasks.

[0062] The aforementioned main robot breaks down its task queue into sub-task queues, specifically categorizing tasks into different types based on their nature and function. For example, in the assembly of new energy streetlights, tasks can be divided into bracket installation, lamp installation, and electrical connection tasks. Tasks within the lamp assembly functional domain can be further broken down by type into sub-tasks such as LED chip mounting, optical lens integration, and lamp housing encapsulation. The main robot first performs the LED chip mounting task, then the optical lens integration task, and finally the lamp housing encapsulation task.

[0063] The auxiliary robot takes over the migration task, obtains the number of migration tasks, adds it to the current task queue number of the auxiliary robot to get the actual number of tasks executed by the auxiliary robot, and compares it with the predefined number of parallel tasks permitted by the auxiliary robot. If the actual number of tasks executed by the auxiliary robot is less than or equal to the number of parallel tasks permitted by the auxiliary robot, the auxiliary robot confirms the takeover of the migration task and maintains the auxiliary robot operation. If the actual number of tasks executed by the auxiliary robot is greater than the number of parallel tasks permitted by the auxiliary robot, the next idle unit in the same attribute operation domain is followed, and the second auxiliary robot performs the operation. The number of tasks executed by the second auxiliary robot is the difference between the actual number of tasks executed by the auxiliary robot and the number of parallel tasks permitted by the auxiliary robot.

[0064] The main robot monitors the status of the auxiliary robot in real time, obtains the real-time motor speed of the auxiliary robot, counts the cumulative time during which the real-time motor speed of the auxiliary robot is less than the preset motor reference speed, and compares it with the predefined cumulative threshold time. If the cumulative time is less than or equal to the cumulative threshold time, the main robot continues to monitor the auxiliary robot. If the cumulative time is greater than the cumulative threshold time, the main robot issues a compensation command.

[0065] Compensation commands typically include adjusting the auxiliary robot's operating speed, increasing its task priority, and optimizing its motion trajectory. In this embodiment, the main robot detects that the auxiliary robot's real-time motor speed is lower than a preset reference speed, and the cumulative duration exceeds a predefined cumulative threshold. The main robot sends a compensation command to the auxiliary robot, for example, "Increase the current operating speed from 50mm / s to 60mm / s." Upon receiving the command, the auxiliary robot adjusts its motor drive parameters, increasing the motor speed to improve its operating speed.

[0066] These instructions are designed to help the auxiliary robot return to normal working condition, ensuring that tasks are completed on time and to a high standard. Through compensation instructions, the main robot can adjust the auxiliary robot's operating mode in a timely manner to cope with potential anomalies. This helps ensure that the auxiliary robot completes tasks according to the established work plan, improving the reliability and stability of the entire production process and ensuring the smooth achievement of production goals.

[0067] By employing more detailed and flexible task breakdown and allocation methods, tasks can be rationally assigned to different auxiliary robots based on their characteristics and capabilities, improving the accuracy and rationality of task allocation. Selecting appropriate auxiliary robots and assigning migration tasks ensures a more balanced workload among robots, preventing some robots from being overloaded while others remain idle, thus improving the overall production line's efficiency and resource utilization. The main robot can flexibly select and assign tasks to auxiliary robots based on real-time production conditions and their status, enhancing the system's dynamic adaptability to the production process.

[0068] Furthermore, the central control port coordinates and adjusts the functional domain tasks to be assembled for each assembly robot. The specific analysis process is as follows:

[0069] Based on the task dependency matrix, the load status of the assembly robot corresponding to the current functional domain task to be assembled is obtained and recorded as the current load status of the assembly robot. The load status can be extracted from the central control port. If the current load status of the assembly robot is a high-load unit, the current functional domain task is divided into a main task queue and a slave task queue according to the task dependency matrix. The slave task queue is recorded as a migration task. The slave robot takes over the migration task, and the main task queue is automatically merged into the task queue of the main robot to form a master-slave collaboration. At the same time, the real-time load evaluation value of the main robot is extracted and mapped to the adapted production cycle of the main robot. The real-time production cycle of the main robot is then adjusted to the adapted production cycle of the main robot.

[0070] The division into main task queues and sub-task queues is specifically based on the task dependency matrix. If a subtask meets any of the following conditions, it is assigned to the main task queue:

[0071] 1) Strongly dependent output: is directly dependent on by two or more subsequent tasks (the number of non-zero elements in the column direction of the matrix is ​​greater than or equal to two).

[0072] 2) Zero-buffered tasks: The allowed delay time is less than or equal to a preset threshold (e.g., greater than or equal to 5 seconds).

[0073] Example:

[0074] LED chip mounting → integration with optical lenses and encapsulation → main task.

[0075] Bracket fixing → solar panel installation, overall machine assembly depends on → main task.

[0076] Lamp housing encapsulation → only dependent on the whole assembly and allowed a delay of 30 seconds → from the task.

[0077] The above mapping yields the adapted production cycle time of the main robot. Specifically, it involves extracting a mapping set between the real-time load assessment value and the adapted production cycle time from the robot management information database, and then inputting the real-time load assessment value of the main robot into the mapping set to obtain the adapted production cycle time of the main robot.

[0078] If the current load state of the assembly robot is an idle unit, then the current assembly robot will execute the task of the functional domain to be assembled.

[0079] Repeat the above process for executing the tasks of the functional domains to be assembled until the coordinated adjustment of all functional domain tasks is completed.

[0080] After completing the task in the functional domain to be assembled, the assembly robot sends the task execution data back to the central control port to determine the execution quality indicators of the assembly robot and to report the current execution status of the assembly robot.

[0081] In this embodiment, a more refined task division method better meets the needs of different tasks, improving the accuracy and flexibility of task management. It helps distinguish between critical tasks and relatively minor tasks, making task management more targeted. Task division is based on resource requirements, rationally allocating resources and avoiding resource contention or idleness. Primary tasks receive sufficient resource guarantees, while remaining resources are flexibly allocated from secondary tasks, improving resource utilization and reducing costs. Prioritizing the execution of primary tasks ensures the smooth completion of critical tasks, guaranteeing a smooth production process and reducing production stoppages caused by delays in critical tasks. Rational allocation of task queues fully utilizes idle resources, reducing resource waste and improving overall production efficiency.

[0082] Specifically, the current execution status of the assembly robot is fed back, and the specific analysis process is as follows:

[0083] The execution quality indicators of the assembly robot are compared with predefined execution quality reference indicators. If the execution quality indicators of the assembly robot are greater than or equal to the execution quality reference indicators, the coordinated adjustment is maintained, and the execution status of the assembly robot is fed back as normal execution.

[0084] If the execution quality index of the assembly robot is less than the execution quality reference index, the current execution status of the assembly robot is reported as abnormal execution. At the same time, the robot self-calibration program is triggered. The joint zero point is recalibrated by laser positioning landmark. The difference between the execution quality index of the assembly robot and the execution quality reference index is processed to obtain the execution quality deviation of the assembly robot. The adaptive scanning frequency of the laser sensor is mapped to obtain the adaptive scanning frequency of the laser sensor. The robot self-calibration program is configured according to the adaptive scanning frequency of the laser sensor.

[0085] The above mapping yields the adaptive scanning frequency of the laser sensor. The mapping set between the execution quality deviation and the adaptive scanning frequency is extracted from the auxiliary robot management information database. The execution quality deviation of the current assembly robot is then incorporated into the mapping set to obtain the adaptive scanning frequency of the laser sensor.

[0086] By monitoring and adjusting the assembly robot's execution status in real time, the system ensures that the robot always operates at its optimal state, reducing production interruptions and rework caused by quality issues. This helps improve the efficiency of the entire production line. A dynamic adjustment mechanism automatically adjusts detection and calibration parameters based on the current execution quality deviation of the assembly robot. This flexibility allows the system to adapt to different production conditions and task requirements. Through laser positioning and joint zero-point recalibration, the system can quickly adapt to changes in robot position or new assembly tasks, improving the system's adaptability and flexibility.

[0087] Furthermore, the execution quality indicators of the assembly robot are determined, and the specific analysis process is as follows:

[0088] Extract the task execution data of the assembly robot, including the overlap of the end effector's movement trajectory, the torque fluctuation variance of the assembly robot, and the standard deviation of the acceleration of each joint of the assembly robot within the task execution cycle. The torque fluctuation variance is monitored by the six-dimensional force sensor built into the assembly robot, and the acceleration of each joint is monitored by the joint encoder.

[0089] The actual trajectory point spatial coordinate set is generated by capturing optical marker points on the end effector in real time using a high-precision visual positioning device. It is then compared one by one with the preset trajectory point spatial coordinate set. Using the three-dimensional Euclidean distance formula between two points in space, the actual trajectory point spatial coordinate set is statistically analyzed. The average of each distance between the actual trajectory point spatial coordinate set and the preset trajectory point spatial coordinate set is then processed by reciprocal, and the overlap of the end effector movement trajectory of the assembly robot is obtained.

[0090] The overlap of the end effector's movement trajectory, the torque fluctuation variance of the assembly robot, and the standard deviation of the acceleration of each joint of the assembly robot are normalized respectively. The normalized results are then weighted and aggregated to obtain the execution quality index of the assembly robot. The specific analysis method is as follows:

[0091]

[0092] In the formula, QL is the execution quality index of the assembly robot, DCT is the overlap degree of the end effector's movement trajectory of the assembly robot, TF is the torque fluctuation variance of the assembly robot, and AC r Let h1 be the standard deviation of the acceleration of the r-th joint of the assembly robot, where r is the joint number, r=[1,R], and R is the total number of joints. h2 is the weight element corresponding to the overlap of the end effector's movement trajectory predefined in the robot management information database, h3 is the weight element corresponding to the torque fluctuation variance predefined in the robot management information database, and h4 is the weight element corresponding to the standard deviation of the joint acceleration predefined in the robot management information database.

[0093] It should be explained that the above normalization process is specifically carried out by first normalizing the numerical range of the end effector's trajectory overlap, torque fluctuation variance, and joint acceleration standard deviation (mapping the values ​​of each physical quantity to a unified interval to eliminate differences in magnitude), and then normalizing the dimensions (removing units through ratio operations of their own ranges to achieve dimensionlessness). This transforms physical quantities with inconsistent units and numerical scales into dimensionless forms with consistent numerical ranges.

[0094] It should be explained that the standard deviation of joint acceleration mentioned above refers to the dispersion of joint acceleration data at each task execution time point. Each task execution time point is a number of time points averaged across the task execution cycle.

[0095] The weight elements corresponding to the end effector trajectory overlap, torque fluctuation variance, and joint acceleration standard deviation are all extracted from the robot management information database. The mapping relationship can be one-to-one or many-to-one. For example, the end effector trajectory overlap, torque fluctuation variance, and joint acceleration standard deviation are mapped to the preset weight elements corresponding to the end effector trajectory overlap, torque fluctuation variance, and joint acceleration standard deviation in the robot management information database to form a mapping set. The real-time end effector trajectory overlap, torque fluctuation variance, and joint acceleration standard deviation are then substituted into the mapping set to obtain the weight elements corresponding to the end effector trajectory overlap, torque fluctuation variance, and joint acceleration standard deviation.

[0096] In this embodiment, multivariate analysis is performed on the end effector trajectory overlap, torque fluctuation variance, and joint acceleration standard deviation. Specifically, the correlation between these parameters is considered. Low end effector trajectory overlap is usually accompanied by increased torque fluctuation variance. This is because trajectory deviation can lead to abnormal forces on the robot's joints, resulting in frequent and increased torque fluctuations, which significantly reduces the robot's performance. Large torque fluctuation variance can cause joint acceleration fluctuations, increasing the joint acceleration standard deviation and thus affecting the end effector's motion stability and trajectory accuracy. Large joint acceleration standard deviations also lead to decreased end effector trajectory overlap and increased torque fluctuation variance, as acceleration fluctuations cause unstable end effector motion, increasing trajectory deviation and torque fluctuation. In short, these parameters interact and jointly determine the robot's assembly quality and production efficiency.

[0097] Specifically, the main robot's allocation of tasks to the auxiliary robot also includes:

[0098] Obtain the real-time task energy consumption value of the auxiliary robot, which can be extracted from the auxiliary robot's execution log. Compare the auxiliary robot's real-time task energy consumption value with a predefined task energy consumption standard threshold. If the auxiliary robot's real-time task energy consumption value is less than or equal to the task energy consumption standard threshold, the main robot assigns tasks to the auxiliary robot. If the auxiliary robot's real-time task energy consumption value is greater than the task energy consumption standard threshold, the difference between the auxiliary robot's real-time task energy consumption value and the task energy consumption standard threshold is calculated to obtain the auxiliary robot's real-time task energy consumption deviation.

[0099] The real-time task energy consumption deviation of the auxiliary robot is compared with the predefined task energy consumption adaptation deviation. If the real-time task energy consumption deviation of the auxiliary robot is less than or equal to the task energy consumption adaptation deviation, the main robot assigns the auxiliary robot the task. If the real-time task energy consumption deviation of the auxiliary robot is greater than the task energy consumption adaptation deviation, a backup robot is called to jointly execute the migration task with the auxiliary robot, forming parallel cooperation with the main robot.

[0100] By comparing the real-time task energy consumption of the auxiliary robot with the standard task energy consumption threshold, tasks are assigned to robots with lower energy consumption and higher efficiency, thereby optimizing task allocation and improving energy utilization efficiency. Through reasonable task allocation, each robot can operate at an appropriate energy consumption level, improving production efficiency while ensuring assembly quality. Dynamically adjusting task allocation based on real-time task energy consumption values ​​allows the central control port to better adapt to changes in production tasks and robot states, enhancing its ability to handle different working conditions. Through master-slave or parallel collaboration modes, the cooperation methods between robots can be flexibly adjusted, improving collaboration efficiency and quality.

[0101] The above description is merely an example and illustration of the structure of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the structure of the invention or exceed the scope defined by the present invention, they should all fall within the protection scope of the present invention.

Claims

1. A multi-robot collaborative control method based on data from a new energy street light production line, characterized in that, include: The central control port receives the overall assembly task of the production line lights, decomposes the task, obtains the task of each functional domain to be assembled, and generates a pre-call signal to the signal receiving port of each assembly robot in the production line. After receiving the pre-call signal, the signal receiving port obtains the real-time execution data of each assembly robot, determines the real-time load assessment value of each assembly robot, feeds back the real-time load status data of each assembly robot to the central control port, and determines whether to perform coordinated adjustment of the assembly robots. Based on real-time load status data, the central control port coordinates and adjusts the assembly functional domain tasks of each assembly robot. Real-time monitoring of the current task execution status of each assembly robot, and feedback on the execution status of each assembly robot; The specific determination process for whether to perform collaborative adjustments to the assembly robot is as follows: The real-time load assessment value of each assembly robot is compared with the predefined load assessment allowable threshold. If the real-time load assessment value of an assembly robot is less than or equal to the load assessment allowable threshold, no coordinated adjustment of the assembly robots is required. The assembly robot is marked as an idle unit and the idle unit execution status data is fed back to the central control port. If the real-time load assessment value of an assembly robot is greater than the load assessment allowable threshold, then the assembly robot needs to be coordinated and adjusted. The assembly robot is marked as a high-load unit, and the execution status data of the high-load unit is fed back to the central control port. At the same time, the difference between the real-time load assessment value of the assembly robot and the load assessment allowable threshold is processed to obtain the load deviation of the assembly robot, and the overflow load status of the assembly robot is determined and coordinated adjustment is performed. The process of determining the overflow load state of the assembly robot and making coordinated adjustments is as follows: The load deviation of the assembly robot is compared with a predefined load deviation threshold. If the load deviation of the assembly robot is less than or equal to the load deviation threshold, the overflow load state of the assembly robot is determined to be overloaded, and the next round of collaborative allocation is triggered first to maintain the current task backlog queue of the assembly robot. If the load deviation of the assembly robot is greater than the load deviation threshold, the overflow load state of the assembly robot is determined to be overloaded, and master-slave collaborative adjustment is initiated. The high-load unit is automatically upgraded to master robot and obtains task scheduling rights. At the same time, the real-time load evaluation value of each idle unit in the same attribute work domain is extracted, and the idle unit with the smallest real-time load evaluation value is recorded as the auxiliary robot. The central control port broadcasts the master-slave relationship change instruction to the auxiliary robot. Based on the task scheduling authority, the main robot assigns tasks to the auxiliary robots.

2. The multi-robot collaborative control method based on new energy street light production line data according to claim 1, characterized in that: The task decomposition and analysis process is as follows: The task decomposition includes primary decomposition and secondary decomposition; The first-level decomposition specifically involves dividing the overall street light assembly task into several production functional domains based on functional characteristics, including the lamp assembly functional domain, the electronic control system assembly functional domain, the solar module installation functional domain, and the complete machine assembly functional domain, with each functional domain corresponding to an independent robot operation unit. The second-level decomposition specifically involves: arranging the tasks of each functional domain to be assembled in the order of execution, ensuring that each task satisfies a preset task dependency matrix, generating a queue of sub-tasks with time-series labels, and allocating each production functional domain according to its spatial location, generating a queue of sub-tasks with spatial labels.

3. The multi-robot collaborative control method based on new energy street light production line data according to claim 1, characterized in that: The specific process for determining the real-time load assessment value of each assembly robot is as follows: The real-time execution data of each assembly robot includes the real-time motor operating load of each assembly robot, the real-time production cycle lag deviation of each assembly robot, the real-time task queue backlog of each assembly robot, and the real-time end effector occupancy rate of each assembly robot. The real-time motor operating load, real-time production cycle lag deviation, real-time task queue backlog, and real-time end effector occupancy rate of each assembly robot are normalized separately. The normalization results are then weighted and aggregated to obtain the real-time load evaluation value of each assembly robot.

4. The multi-robot collaborative control method based on new energy street light production line data according to claim 1, characterized in that: The process of assigning tasks to the auxiliary robot by the main robot is as follows: The main robot splits its own task backlog into sub-backlog tasks, obtains the waiting time of each sub-backlog task, and records several sub-backlog tasks whose waiting time exceeds the preset waiting time limit as migration tasks. The auxiliary robot takes over the migration task, obtains the number of migration tasks, adds it to the current task queue number of the auxiliary robot to get the actual number of tasks executed by the auxiliary robot, and compares it with the predefined number of parallel tasks permitted by the auxiliary robot. If the actual number of tasks executed by the auxiliary robot is less than or equal to the number of parallel tasks permitted by the auxiliary robot, the auxiliary robot confirms the takeover of the migration task and maintains the auxiliary robot operation. If the actual number of tasks executed by the auxiliary robot is greater than the number of parallel tasks permitted by the auxiliary robot, the next idle unit in the same attribute operation domain is followed, and the second auxiliary robot performs the operation. The number of tasks executed by the second auxiliary robot is the difference between the actual number of tasks executed by the auxiliary robot and the number of parallel tasks permitted by the auxiliary robot. The main robot monitors the status of the auxiliary robot in real time, obtains the real-time motor speed of the auxiliary robot, counts the cumulative time during which the real-time motor speed of the auxiliary robot is less than the preset motor reference speed, and compares it with the predefined cumulative threshold time. If the cumulative time is less than or equal to the cumulative threshold time, the main robot continues to monitor the auxiliary robot. If the cumulative time is greater than the cumulative threshold time, the main robot issues a compensation command.

5. The multi-robot collaborative control method based on new energy street light production line data according to claim 1, characterized in that: The central control port coordinates the assembly tasks of each assembly robot in the functional domain to be assembled. The specific analysis process is as follows: Based on the task dependency matrix, the load status of the assembly robot corresponding to the current functional domain task to be assembled is obtained and recorded as the current load status of the assembly robot. If the current load status of the assembly robot is a high load unit, the current functional domain task is divided into a main task queue and a secondary task queue according to the task dependency matrix. The secondary task queue is recorded as a migration task. The auxiliary robot takes over the migration task, and the main task queue is automatically merged into the task queue of the main robot to form a master-slave collaboration. At the same time, the real-time load evaluation value of the main robot is extracted and mapped to the adapted production cycle of the main robot. The real-time production cycle of the main robot is then adjusted to the adapted production cycle of the main robot. If the current load state of the assembly robot is an idle unit, then the current assembly robot will execute the task of the functional domain to be assembled. Repeat the above process for executing the tasks of the functional domains to be assembled until the coordination and adjustment of all functional domain tasks are completed. After completing the task in the functional domain to be assembled, the assembly robot sends the task execution data back to the central control port to determine the execution quality indicators of the assembly robot and to report the current execution status of the assembly robot.

6. The multi-robot collaborative control method based on new energy street light production line data according to claim 5, characterized in that: The specific analysis process for providing feedback on the current execution status of the assembly robot is as follows: The execution quality indicators of the assembly robot are compared with the predefined execution quality reference indicators. If the execution quality indicators of the assembly robot are greater than or equal to the execution quality reference indicators, the coordinated adjustment is maintained and the execution status of the assembly robot is fed back as normal execution. If the execution quality index of the assembly robot is less than the execution quality reference index, the current execution status of the assembly robot is reported as abnormal execution. At the same time, the robot self-calibration program is triggered. The joint zero point is recalibrated by laser positioning landmark. The difference between the execution quality index of the assembly robot and the execution quality reference index is processed to obtain the execution quality deviation of the assembly robot. The adaptive scanning frequency of the laser sensor is mapped to obtain the adaptive scanning frequency of the laser sensor. The robot self-calibration program is configured according to the adaptive scanning frequency of the laser sensor.

7. The multi-robot collaborative control method based on new energy street light production line data according to claim 5, characterized in that: The specific analysis process for determining the performance quality indicators of the assembly robot is as follows: Extract the task execution data of the assembly robot, including the overlap of the end effector's movement trajectory, the torque fluctuation variance of the assembly robot, and the standard deviation of the acceleration of each joint of the assembly robot. The actual trajectory point spatial coordinate set is generated by capturing optical marker points on the end effector in real time using a high-precision visual positioning device. It is then compared one by one with the preset trajectory point spatial coordinate set. Using the three-dimensional Euclidean distance formula between two points in space, the actual trajectory point spatial coordinate set is statistically analyzed. The average of each distance between the actual trajectory point spatial coordinate set and the preset trajectory point spatial coordinate set is then processed by reciprocal, and the overlap of the end effector movement trajectory of the assembly robot is obtained. The overlap of the end effector's movement trajectory, the torque fluctuation variance of the assembly robot, and the standard deviation of the acceleration of each joint of the assembly robot are normalized respectively. The normalization results are then weighted and aggregated to obtain the execution quality index of the assembly robot.

8. The multi-robot collaborative control method based on new energy street light production line data according to claim 4, characterized in that: The process of assigning tasks to the auxiliary robot by the main robot also includes: The real-time task energy consumption value of the auxiliary robot is obtained and compared with the predefined task energy consumption standard threshold. If the real-time task energy consumption value of the auxiliary robot is less than or equal to the task energy consumption standard threshold, the main robot assigns the auxiliary robot the task. If the real-time task energy consumption value of the auxiliary robot is greater than the task energy consumption standard threshold, the difference between the real-time task energy consumption value of the auxiliary robot and the task energy consumption standard threshold is processed to obtain the real-time task energy consumption deviation of the auxiliary robot. The real-time task energy consumption deviation of the auxiliary robot is compared with the predefined task energy consumption adaptation deviation. If the real-time task energy consumption deviation of the auxiliary robot is less than or equal to the task energy consumption adaptation deviation, the main robot assigns the auxiliary robot the task. If the real-time task energy consumption deviation of the auxiliary robot is greater than the task energy consumption adaptation deviation, a backup robot is called to jointly execute the migration task with the auxiliary robot, forming parallel cooperation with the main robot.

Images