Method, system, device and medium for fair pace control of robot collaborative production line

Abstract

The application discloses a robot collaborative production line fair beat control method, system, equipment and medium, the method acquires production line equipment group state information and work task data by a state monitoring unit; a virtual balanced production line reference control model is established to obtain the reference fair beat capacity of each work task, and the actual fair beat capacity is calculated based on the current control strategy; the deviation between the two is used to determine the fair keeping proportion, the illegal task set and the tail control index of the synchronization batch; the risk control amount is calculated based on the entrance station competition degree, the exit station competition degree, the candidate process path overlapping degree and the normalized bottleneck load, and the process path execution state selection, the robot task reassignment, the buffer release, the conveying line shunt or the bottleneck station accommodation adjustment action is generated accordingly; the control strategy is updated according to the operation state feedback by a feedback verification unit. The application realizes the closed-loop control of the fair beat and the production capacity balance of the industrial robot collaborative production line.

Description

Technical Field

[0001] This invention relates to the field of collaborative production line monitoring and control technology, and in particular to a method, system, equipment and medium for fair cycle control of robot collaborative production lines. Background Technology

[0002] Existing smart factories typically include manufacturing execution systems, programmable logic controllers (PLCs), robot controllers, warehouse scheduling systems, and conveyor controllers. These controllers can respectively implement equipment start / stop, workstation dispatching, path handling, buffer release, or local cycle time optimization. However, in practical applications, the following technical shortcomings still exist: First, there is a lack of a unified reference input mechanism. Existing strategies typically compare equipment load or total capacity directly, making it difficult to quantify the structural degradation of the cycle time capability of individual work tasks caused by real production line bottlenecks, and thus failing to generate comparable feedback deviations.

[0003] Second, there is insufficient protection for tail tasks. A small number of tasks in hotspot entry points, hotspot target workstations, shared robots, and synchronous order batches may remain in a low service takt time state for extended periods, and existing solutions lack targeted protection mechanisms for tail tasks.

[0004] Third, there is a lack of an adjustable mechanism between fairness and capacity. Different orders have different preferences for fairness intensity, capacity, and simultaneous batch completion time, and fixed control targets cannot adapt to diverse production needs.

[0005] Fourth, there is a lack of budget constraints on changes in the execution status of work processes in online scenarios. Frequent migration of work tasks increases the risks associated with rule issuance, state transitions, model changes, and fault recovery, and existing solutions lack effective limitations in addressing these issues.

[0006] Fifth, local search failures are easily misjudged as infeasibility. Failure to meet the fairness threshold may be due to insufficient candidate process paths, insufficient repair rounds, or insufficient reconfiguration budget, rather than necessarily an unsolvable problem in the physical production line itself. Existing solutions lack interpretable failure state outputs.

[0007] In summary, existing solutions typically focus on local scheduling under the execution status of equipment utilization, total capacity, or a given process path, and lack a closed-loop control solution that integrates production line equipment operation status monitoring, reference control models, feedback deviations, adjustment control quantities, budget constraint execution, and online feedback verification into a general control or adjustment system. Summary of the Invention

[0008] Therefore, the purpose of this invention is to provide a method, system, device, and medium for fair cycle control in robot collaborative production lines to solve the above-mentioned problems.

[0009] To achieve the above-mentioned objectives, this invention provides a fair cycle time control method for robot collaborative production lines, applicable to general control or regulation systems. The general control or regulation system includes a controller, a status monitoring unit, a control command issuing unit, and a feedback verification unit, comprising the following steps: Step S1: The status monitoring unit monitors or tests the industrial automation production line equipment group and collects information such as workstation topology, equipment or workstation capacity, workstation status, buffer status, robot posture, fixture status, work-in-process position, and task status.

[0010] Specifically, the controller or control agent monitors and tests the controlled industrial automation production line equipment group through the status monitoring unit via the industrial site management interface, industrial robot telemetry, programmable logic controller status interface, industrial robot collaborative operation connection status, cache statistics, equipment or workstation utilization statistics, or a pre-configured workstation topology database. The collected data includes: workstation topology relationships, equipment or workstation capacity, effective cycle time capacity of workstations, equipment or workstation availability and fault status; entry workstation, exit workstation, remaining processing volume, setup time, duration, equipment interface type, order priority, and synchronous order batches for work tasks; entry workstation competition level, exit workstation competition level, equipment or workstation utilization rate, cache occupancy, congestion alarm ratio, and process path control from the previous control cycle. For fields that cannot be obtained in real time, configured values, sliding window averages, the most recent valid value, or values ​​reported by the business side can be used. The collection period can be stage-level, millisecond-level, second-level, or event-level triggered by the business controller, specifically determined by the production line scale and equipment sampling capabilities.

[0011] Step S2: Generate a set of candidate process path execution states that can be executed by the controlled industrial equipment for the job task.

[0012] Specifically, for each entrance and exit pair ,in, t represents the entry station, and t represents the exit station. The controller uses station topology search, pre-calculated process path library, segmented process path set, existing multi-path execution strategy, next station set, or candidate process paths provided by the manufacturing execution system. Candidate execution states are generated. Candidate process paths can be restricted to process paths that pass through different intermediate industrial robots, meet equipment or workstation reliability constraints, meet process path length constraints, or avoid faulty equipment or workstations. Each job task selects only one process path from the candidate set of the corresponding inlet-outlet pair, forming an indivisible single-process path execution control state. The controller does not require the same job task to be finely divided among multiple candidate process paths.

[0013] Step S3: Establish a reference control model to characterize inlet-side constraints, outlet-side constraints, and key process constraints, and calculate the reference fair cycle time capability of each task based on the reference control model.

[0014] Specifically, the controller establishes a virtual equilibrium production line model VBL(G), where G represents the production line topology, including workstation nodes and the connections between them. The access capacity of the entry and exit workstations is used as constraints, ignoring the shared constraints of intermediate workstations, but retaining the access capacity, critical process sequence, and safety interlock constraints on the entry and exit sides. The controller performs a max-min fairness calculation on this reference feasible region to obtain the reference fair takt time capability for each task. This reference fair takt time capability is the takt time capability of each task obtained through max-min fairness calculation under the reference control model, denoted as [missing information]. This reference cycle time capability serves as a reference value for regulation and control. Its purpose is to normalize the cycle time capability loss caused by internal bottlenecks, ensuring comparability of fair maintenance ratios across different workstation topologies, different combinations of workloads, and different control cycles.

[0015] Step S4: Based on the single-process path execution status under the current control strategy and the capacity constraints of the equipment or workstation, calculate the actual fair cycle time capability of each task.

[0016] Specifically, the controller constructs a workstation-task association matrix A based on the process path execution state r corresponding to the current control strategy, where... Indicates the task / task Use equipment or workstations Otherwise, it is 0. For workstation-task association matrix elements, For homework assignments The beat ability variable, For equipment or workstation The capacity for each device or workstation Apply capacity constraints And under this constraint, calculations are performed using water level filling allocation, linear programming, hierarchical bottleneck elimination, or a max-min fairness solver. The This represents the actual fair cycle time capability of task f under the current process path execution state r corresponding to the control strategy.

[0017] Step S5: Based on the deviation between the actual fair beat capability and the reference fair beat capability, determine the fairness maintenance ratio, the set of non-compliant tasks, and the synchronization batch tail control index to form a feedback deviation amount.

[0018] Specifically, the fairness maintenance ratio is defined as

[0019] in, Let $f$ be the reference fair tick capacity, i.e., the reference fair tick capacity. Given a fairness threshold... Fairness and violations are collectively referred to as The controller also calculates the minimum fairness maintenance ratio. Low percentile fairness maintains proportion ,in, To maintain the minimum proportion fairly across all tasks. To ensure fairness across all tasks, a low quantile, such as the 10th percentile, is maintained to prevent a single abnormal task from completely dominating the evaluation results. The synchronous batch tail control index is determined by the maximum of the ratios of the remaining processing capacity of each task in the synchronous order batch to the actual fair cycle time capability. The synchronous batch tail control index is defined as

[0020] in For homework assignments The remaining processing quantity, the synchronous order batch refers to the set of order tasks with synchronous completion requirements, and this indicator is used as the monitoring or test output of the general control or regulation system to characterize the tail waiting risk of the controlled production batch.

[0021] Step S6: Calculate the risk control quantity of the controlled task based on the competition degree of the entry station, the competition degree of the exit station, the overlap of candidate process paths, the normalized bottleneck load, and the importance of order synchronization.

[0022] Specifically, calculate the risk control amount for each task. One implementation method is as follows: ,in Competition for entry-level workstations. To assess the competitiveness of export workstations. The overlap of candidate process paths. To normalize the bottleneck load, Importance and weight of order synchronization Configurable. The weights correspond to the entry station competition degree, exit station competition degree, candidate process path overlap degree, normalized bottleneck load, and order synchronization importance, respectively. The entry station competition degree represents the degree to which a task competes with other tasks for the same entry station, and can be determined based on at least one of the following: the number of tasks queuing for the same entry station, waiting time, and task priority. The exit station competition degree represents the degree to which a task competes with other tasks for the same exit station, and can be determined based on at least one of the following: the number of tasks queuing for the same exit station, waiting time, and task priority. The candidate process path overlap degree represents the degree to which candidate process paths between tasks share intermediate equipment or stations, and can be determined based on the ratio of the number of intermediate stations shared by two task candidate process paths to the total number of intermediate stations traversed by the candidate process path. The risk control quantity is obtained by weighting at least two of the following: entry station competition degree, exit station competition degree, candidate process path overlap degree, normalized bottleneck load, and order synchronization importance, and is used to determine the processing order of tasks and construct an inter-task conflict graph. In another implementation, the risk control quantity can also be adjusted based on the equipment health status to modify the weighted result. For example, when the equipment health status is below a set threshold, the risk control quantity of the corresponding task can be increased.

[0023] Step S7: Based on the risk control amount and according to the hierarchical feedback adjustment target, generate adjustment actions, including process path execution state selection, robot task reallocation, buffer release, conveyor line diversion, workpiece reordering, or bottleneck station clearance.

[0024] Specifically, according to The job tasks are processed in descending order. For the current job tasks to be assigned, the conflict load, normalized bottleneck load, process path length, control rule change cost, and synchronous order batch impact of each candidate process path are calculated, and the candidate process path with the lowest comprehensive score is selected. The comprehensive score is obtained by weighting the conflict load, normalized bottleneck load, process path length, control rule change cost, and synchronous order batch impact; wherein, the greater the conflict load, normalized bottleneck load, process path length, control rule change cost, and synchronous order batch impact, the higher the comprehensive score, and the candidate process path with the lowest comprehensive score is selected for execution. The conflict load represents the degree of conflict between the candidate process path and the assigned task path, and can be determined, for example, based on the number of devices or workstations shared by the candidate process path and the assigned task path; the process path length represents the number of workstations traversed by the candidate process path; the control rule change cost represents the cost of changing control rules to switch to the candidate process path, and can be determined, for example, based on at least one of the number of control rule changes, the number of fixture changes, or the number of job procedure changes; the synchronous order batch impact represents the impact of the candidate process path on the tail control indicators of the synchronous order batch, and can be determined, for example, based on the change in the tail control indicators of the synchronous batch corresponding to the candidate process path. The overlap relationship of candidate process paths between job tasks is written into the conflict diagram. If two job tasks share more than a set proportion of intermediate workstations, share high-load equipment or workstations, or cause one job task to fall below the fairness threshold, then... Establish conflict borders in China.

[0025] In one implementation, the set of work tasks is divided into a protection task set and a core task set. The protection task set consists of work tasks for which the fairness maintenance ratio must be prioritized to be no less than a preset fairness threshold. The core task set consists of work tasks for which the total production line capacity must be prioritized to be increased and the bottleneck load reduced. The protection task set generates adjustment actions based on the fairness takt time maintenance target, while the core task set generates adjustment actions based on the capacity improvement target. The fairness takt time maintenance target refers to ensuring that the fairness maintenance ratio of the work tasks in the protection task set is no less than the preset fairness threshold. The capacity improvement target refers to increasing the total production line capacity corresponding to the core task set and reducing the bottleneck load. Protection tasks are identified based on the inlet / outlet bipartite graph, inlet / outlet competition, synchronous order batches of work tasks, and the estimated fairness maintenance ratio. The estimated fairness maintenance ratio is the fairness maintenance ratio calculated based on the previous control cycle or the current initial allocation result. Protection tasks prioritize candidate process paths that can increase the fairness maintenance ratio; core tasks prioritize candidate process paths that can increase total capacity and reduce bottleneck load. When a fairness violation task exists, a local adjustment action is selected according to the hierarchical feedback adjustment target.

[0026] In another implementation, within a dynamic control cycle, the path execution state of continuous tasks from the previous control cycle is inherited. The continuous tasks are those that existed in the previous control cycle and are still being processed in the current control cycle; the newly arriving tasks are those newly added to the current control cycle. For newly arriving tasks, the equipment or workstation reference load is calculated based on the fair reference cycle time capacity of each task under the balanced production line reference model. The summation object is the currently used equipment or workstation. The task at hand. For newly arrived task assignments. Candidate process paths Calculate the normalized bottleneck load

[0027] And select The smallest candidate process path. When the fairness violation task is detected, only on-budget control migrations are performed on continuous tasks, and the number of migrations does not exceed the preset reconfiguration budget.

[0028] Step S8: Within the preset repair budget or reconfiguration budget, the control command issuing unit issues robot operation program selection instructions, programmable logic controller variables, conveyor line diversion instructions, buffer release instructions, fixture switching instructions, or automatic guided transport equipment scheduling instructions to the controlled industrial equipment.

[0029] Specifically, the controller converts the mapping of job tasks to process paths into one or more control configurations, including workpiece batch matching rules, workpiece tags, next station groups, segmented process path lists, programmable logic controller cache binding, fixture or station selection strategies, or equivalent job programs or PLC variables. For equipment requiring rule synthesis, the controller first performs a control resource check to confirm that the number of control table entries, priority space, and update window meet the constraints before issuing the update. Here, atomic version switching refers to switching the process path control table as a whole; dual-table switching refers to maintaining the current table and a backup table, switching after the backup table is updated; and phased issuance refers to issuing updates in batches according to job task groups or control areas. Process path execution status updates can employ atomic version switching, dual-table switching, phased issuance, switching by job task group, or incremental updates with rollback flags.

[0030] Step S9: The feedback verification unit updates the control strategy based on the running status feedback after the instruction is issued.

[0031] If the update fails, the device is unreachable, the control table entry exceeds the limit, or the telemetry result is abnormal, the controller will revert to the previous version of the process path control and record this action in the failure log. When the repair rounds reach... Stop when there are no actions to improve or the fairness threshold has been met, and output the process path control table, feasibility flags, and repair statistics. Failure to completely clear within round L. At that time, among them, To ensure fairness in the collection of tasks that violate regulations, Return to the process path control with the fewest violation tasks, the highest tail fairness, and a production capacity not lower than the set lower limit among the visited control states, set the feasible flag to false, and output the violation rate (FailureFraction) and the average number of repair rounds. The repair budget or reconfiguration budget refers to the maximum number of control migrations allowed or the maximum number of control table changes allowed, which can be set according to the production line scale and real-time requirements.

[0032] The present invention also provides a general control or adjustment system for fair cycle control of robot collaborative production lines, including a controller, a state monitoring unit, a reference control model unit, a deviation calculation unit, a risk control quantity calculation unit, an adjustment action generation unit, a budget constraint execution unit, a control command issuance unit, and a feedback verification unit. The status monitoring unit is used to monitor or test the equipment group of industrial automated production line, and collect the workstation topology, equipment or workstation capacity, workstation status, buffer status, robot posture, fixture status, work-in-process position and task status. The controller is used to generate a set of candidate process path execution states that can be executed by the controlled industrial equipment for the job task; The reference control model unit is used to establish a reference control model to characterize inlet-side constraints, outlet-side constraints, and critical process constraints, and to calculate the reference fair cycle time capability of each task; the reference fair cycle time capability is the cycle time capability of each task obtained by max-min fairness calculation under the reference control model. The deviation calculation unit is used to calculate the actual fair cycle time capability based on the single-process path execution status and equipment or workstation capacity constraints under the current control strategy. The actual fair cycle time capability is the cycle time capability of each task obtained through maximum and minimum fairness calculation under the single-process path execution status and equipment or workstation capacity constraints under the current control strategy. It also determines the fairness maintenance ratio, the set of non-compliant tasks, and the synchronous batch tail control index based on the deviation between the actual fair cycle time capability and the reference fair cycle time capability, forming a feedback deviation amount. The fairness maintenance ratio is the ratio of the actual fair cycle time capability to the reference fair cycle time capability. When the fairness maintenance ratio is lower than a preset fairness threshold, the corresponding task is identified as a fairness non-compliant task, and the set of non-compliant tasks is composed of these tasks. The synchronous batch tail control index is determined by the maximum value among the ratios of the remaining processing volume of each task in the synchronous order batch to the actual fair cycle time capability. The risk control quantity calculation unit is used to calculate the risk control quantity of the controlled task based on at least two of the following: entry station competition degree, exit station competition degree, candidate process path overlap degree, normalized bottleneck load, and order synchronization importance. The entry station competition degree indicates the degree to which a task competes with other tasks for the same entry station. The exit station competition degree indicates the degree to which a task competes with other tasks for the same exit station. The candidate process path overlap degree indicates the degree to which candidate process paths among tasks share intermediate equipment or stations. The order synchronization importance indicates the degree to which the task belongs to the synchronous order batch's synchronous completion requirement, or the degree of influence of the task on the completion time of the synchronous order batch. The normalized bottleneck load is determined by the maximum value among the ratios of the sum of the reference load of each device or station in the candidate process path and the reference fair takt capacity of the task to be assigned, and the capacity of the corresponding device or station. The reference load is the sum of the reference fair takt capacities of all tasks currently using that device or station. The adjustment action generation unit is used to generate adjustment actions based on the risk control amount and according to the hierarchical feedback adjustment target. The adjustment actions include process path execution state selection, robot task reallocation, buffer release, conveyor line diversion, workpiece reordering, or bottleneck station clearance. The hierarchical feedback adjustment target refers to the target of selecting local adjustment actions according to the priority order of the number of non-compliant tasks, minimum fairness maintenance ratio, low percentile fairness maintenance ratio, and synchronous batch tail control index. The minimum fairness maintenance ratio is the minimum value among all work task fairness maintenance ratios. The low percentile fairness maintenance ratio is the low percentile value among all work task fairness maintenance ratios. The budget constraint execution unit is used to perform control migration within a preset repair budget or reconfiguration budget; The control command issuing unit is used to issue robot operation program selection instructions, programmable logic controller variables, conveyor line diversion instructions, buffer release instructions, fixture switching instructions or automatic guided transportation equipment scheduling instructions to the controlled industrial equipment. The feedback verification unit is used to update the control strategy based on the feedback of the operating status.

[0033] The present invention also provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the computer program, when executed on the processor, causes the electronic device to perform the above-described method.

[0034] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when run on a processor, causes the processor to perform the above-described method.

[0035] Compared with the prior art, the beneficial effects of the present invention are: First, by using the ratio of actual fair cycle time capability to reference fair cycle time capability, the service degradation caused by intermediate workstation bottlenecks is transformed into a calculable fair maintenance ratio, which facilitates comparison and parameter adjustment by the controller and achieves normalization of the reference input.

[0036] Second, by using risk ranking and conflict graph construction, hotspot entry points, hotspot exit points, internal shared workstations, and synchronous tail tasks are incorporated into the same process path execution status selection mechanism, which effectively reduces the probability of high-risk tasks being squeezed out by capacity targets and improves the fairness of tail tasks.

[0037] Third, by using fair thresholds and hierarchical repair targets, the system prioritizes reducing non-compliant tasks when feasible, and outputs interpretable failure statuses (including feasible flags, proportion of non-compliant tasks, and number of repair rounds) when infeasible or insufficient search is required. This avoids erroneous conclusions of infeasibility and improves the system's diagnosability.

[0038] Fourth, by reconfiguring the budget online to limit the number of job task process path migrations, the operational risks caused by frequent changes in control instructions, state transitions, and control table updates are effectively reduced, and the stability of the system in dynamic scenarios is improved.

[0039] Fifth, by linking task-level fairness with application-stage waiting through synchronous batch tail control indicators, the tail congestion problem in synchronous batch production is effectively improved, reducing industrial robot waiting, buffer congestion, and station idleness. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only preferred embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a schematic diagram of the architecture of a general control or regulation system for fair cycle control of a robot collaborative production line according to the present invention.

[0042] Figure 2 This is a schematic diagram of the robot collaborative production line fair cycle control method provided by the present invention. Detailed Implementation

[0043] The following is in conjunction with the appendix Figure 1-2 The principles and features of the present invention are described herein, and the listed embodiments are only used to explain the present invention and are not intended to limit the scope of the present invention.

[0044] Example 1: Fair cycle time adjustment in industrial robot welding production line This embodiment uses an industrial robot welding production line as an example to explain in detail the implementation process of the method of the present invention.

[0045] See Figure 1 The general control or regulation system architecture of the present invention includes a controller, a state monitoring unit, a reference control model unit, a deviation calculation unit, a risk control quantity calculation unit, a regulation action generation unit, a budget constraint execution unit, a control command issuance unit, and a feedback verification unit. The status monitoring unit is used to monitor or test the equipment group of industrial automated production line, and collect the workstation topology, equipment or workstation capacity, workstation status, buffer status, robot posture, fixture status, work-in-process position and task status. The controller is used to generate a set of candidate process path execution states that can be executed by the controlled industrial equipment for the job task; The reference control model unit is used to establish a reference control model to characterize inlet-side constraints, outlet-side constraints, and critical process constraints, and to calculate the reference fair cycle time capability of each task; the reference fair cycle time capability is the cycle time capability of each task obtained by max-min fairness calculation under the reference control model. The deviation calculation unit is used to calculate the actual fair cycle time capability based on the single-process path execution status and equipment or workstation capacity constraints under the current control strategy. The actual fair cycle time capability is the cycle time capability of each task obtained through maximum and minimum fairness calculation under the single-process path execution status and equipment or workstation capacity constraints under the current control strategy. It also determines the fairness maintenance ratio, the set of non-compliant tasks, and the synchronous batch tail control index based on the deviation between the actual fair cycle time capability and the reference fair cycle time capability, forming a feedback deviation amount. The fairness maintenance ratio is the ratio of the actual fair cycle time capability to the reference fair cycle time capability. When the fairness maintenance ratio is lower than a preset fairness threshold, the corresponding task is identified as a fairness non-compliant task, and the set of non-compliant tasks is composed of these tasks. The synchronous batch tail control index is determined by the maximum value among the ratios of the remaining processing volume of each task in the synchronous order batch to the actual fair cycle time capability. The risk control quantity calculation unit is used to calculate the risk control quantity of the controlled task based on at least two of the following: entry station competition degree, exit station competition degree, candidate process path overlap degree, normalized bottleneck load, and order synchronization importance. The entry station competition degree indicates the degree to which a task competes with other tasks for the same entry station. The exit station competition degree indicates the degree to which a task competes with other tasks for the same exit station. The candidate process path overlap degree indicates the degree to which candidate process paths among tasks share intermediate equipment or stations. The order synchronization importance indicates the degree to which the task belongs to the synchronous order batch's synchronous completion requirement, or the degree of influence of the task on the completion time of the synchronous order batch. The normalized bottleneck load is determined by the maximum value among the ratios of the sum of the reference load of each device or station in the candidate process path and the reference fair takt capacity of the task to be assigned, and the capacity of the corresponding device or station. The reference load is the sum of the reference fair takt capacities of all tasks currently using that device or station. The adjustment action generation unit is used to generate adjustment actions based on the risk control amount and according to the hierarchical feedback adjustment target. The adjustment actions include process path execution state selection, robot task reallocation, buffer release, conveyor line diversion, workpiece reordering, or bottleneck station clearance. The hierarchical feedback adjustment target refers to the target of selecting local adjustment actions according to the priority order of the number of non-compliant tasks, minimum fairness maintenance ratio, low percentile fairness maintenance ratio, and synchronous batch tail control index. The minimum fairness maintenance ratio is the minimum value among all work task fairness maintenance ratios. The low percentile fairness maintenance ratio is the low percentile value among all work task fairness maintenance ratios. The budget constraint execution unit is used to perform control migration within a preset repair budget or reconfiguration budget; The control command issuing unit is used to issue robot operation program selection instructions, programmable logic controller variables, conveyor line diversion instructions, buffer release instructions, fixture switching instructions or automatic guided transportation equipment scheduling instructions to the controlled industrial equipment. The feedback verification unit is used to update the control strategy based on the feedback of the operating status.

[0046] Each unit forms a closed-loop control link with the controlled industrial equipment group.

[0047] A two-tiered industrial automation production line is constructed, comprising 8 entry stations, 8 exit stations, and 3 intermediate industrial robots. In this line, the entry stations handle workpiece loading, the intermediate industrial robots handle welding operations, and the exit stations handle unloading and inspection. The effective capacity of the intermediate stations is 0.25 times the capacity of the edge control equipment or stations; this ratio can be set based on the actual sharing and bottleneck levels of the intermediate stations.

[0048] The controller first generates 240 indivisible job tasks, including hotspot inlet / outlet, inlet station stacking, outlet station stacking, and background jobs. Each job task has different inlet stations, outlet stations, and remaining processing capacity.

[0049] See Figure 2 In the closed-loop control method flow of the present invention shown, in step S1, the status monitoring unit collects the status information of the production line equipment group, including the topology of each workstation, equipment capacity, buffer level and task queue.

[0050] In step S2, the controller generates a set of candidate process paths for each inlet and outlet pair, and each job task is limited to selecting one indivisible process path from the candidate set.

[0051] In step S3, the controller establishes a virtual balanced production line model VBL(G), retaining the access capacity constraints on the inlet and outlet sides, ignoring the intermediate workstation sharing constraints, and obtaining the reference fair cycle time capacity of each job task through max-min fair calculation. .

[0052] In step S4, the controller constructs a workstation-task association matrix based on the current process path control state, applies equipment capacity constraints, and calculates the actual fair cycle time capability of each task. .

[0053] In step S5, the controller calculates the fair maintenance ratio for each job task. Identify tasks that violate fairness rules and calculate the tail control indicators for synchronous batches.

[0054] In step S6, the controller calculates the risk control quantity; in step S7, the controller generates adjustment actions based on the risk control quantity and according to the hierarchical feedback adjustment target, and performs budget screening; in step S8, the control command issuing unit issues control commands to the PLC, robot, or conveyor line; in step S9, the feedback verification unit updates the control strategy based on cycle time, buffer, or alarm feedback.

[0055] Risk control parameters can be weighted by the competition degree of the entry station, the competition degree of the exit station, the overlap of candidate process paths, the normalized bottleneck load, and the importance of order synchronization; the controller further constructs a conflict graph between tasks based on the risk control parameters. If two job tasks share more than a set proportion of intermediate workstations, share high-load equipment or workstations, or cause one job task to fall below the fairness threshold, then... Establish conflict borders in China.

[0056] In steps S6 and S7, the controller employs a bottleneck-aware fairness-maintaining adjustment (BAFR) strategy, processing job tasks in descending order of risk control quantity. Initial process path allocation is performed, and after initial allocation, the fair takt time capability is recalculated. For tasks violating fairness rules, single task migration, process path swapping between two tasks, and yielding actions for high-capacity tasks are attempted sequentially. After completing no more than L rounds of repairs, a process path control table and repair statistics are output.

[0057] This embodiment effectively reduces the waiting time for tasks at the hotspot inlet and outlet through the BAFR strategy, shortening the synchronous batch completion time of the welding production line by approximately 15% to 20% compared to traditional methods.

[0058] Example 2: Capacity Boost Control under Fairness Threshold This embodiment sets a fair threshold based on the production line status and workload of Embodiment 1. =0.4 to 0.6, employing the Fair Protection Beat Adjustment (FGTR) strategy.

[0059] In this embodiment, when executing step S7, the controller first identifies the set of protection tasks based on the inlet / outlet bipartite graph, the inlet / outlet competition degree, and the synchronous order batches of the job tasks. and core task set .

[0060] For protection tasks, priority is given to candidate process paths that can improve the fair retention rate, ensuring that the fair retention rate is not lower than a set threshold. For core tasks, priority is given to candidate process paths that can increase total capacity and reduce bottleneck load, thereby improving overall output.

[0061] After the initial allocation is completed, the controller calculates the set of fair violations. .like If not empty, attempt the following actions in sequence: migrating violating tasks, exchanging conflicting tasks, and yielding core tasks. Recalculate the actual fair tick capacity and synchronization batch completion time for each candidate action. Action acceptance rules are executed in the following order: first reduce the number of violating tasks; then, if the number of violating tasks is the same, increase the minimum fairness maintenance ratio. ;exist Increase the proportion of fair play at the lower percentile when they are the same When the aforementioned indicators are the same, reduce the tail control indicators of the synchronous batch. .

[0062] exist The inside of the wheel was not completely emptied. When the controller returns the process path control with the fewest violation tasks and the highest tail fairness among the control states it has visited, it sets the feasible flag to false and outputs the violation rate FailureFraction.

[0063] This embodiment achieves flexible adjustment between fairness and capacity under different fairness thresholds, and can dynamically adjust the control target according to the priority of different orders, making it suitable for multi-variety mixed-flow production scenarios.

[0064] Example 3: Coordinated Control of Automated Guided Transport Equipment and Buffer Stations This embodiment constructs a production line unit containing 12 entry workstations, 12 exit workstations, and 4 intermediate industrial robots, generating full-to-full production operations and full-to-reduction production operation proxy tasks.

[0065] Set the synchronized order batch size (groupSize=6), synchronized order batch quantity (groupCount=10), and the effective capacity of intermediate workstations to 0.30. Each job task is marked with a synchronized order batch identifier. The controller will synchronize the batch tail control indicators. As the final judgment indicator during control and repair.

[0066] The production line is equipped with automated guided vehicles (AGVs) responsible for material transfer between workstations, as well as buffer workstations for temporary storage of workpieces awaiting processing. The adjustment actions generated by the controller are converted into control commands by the control command issuing unit, including AGV scheduling commands (including handling task allocation, travel path selection, and charging scheduling) and buffer release commands (including buffer release timing and release priority).

[0067] During the control process, when the tail control indicator of a certain synchronous order batch... When the temperature is too high, the controller prioritizes assigning high-priority automated guided transport equipment and buffer release permissions to the task with the largest remaining processing volume in the batch, thereby shortening its waiting time.

[0068] This embodiment verifies the applicability of the method of the present invention in complex material transfer scenarios that include automated guided transport equipment and buffer workstations, effectively improving the tail waiting problem of synchronous order batches, and reducing the maximum deviation of batch completion time by about 25%.

[0069] Example 4: Dynamic order insertion and bounded reconfiguration control This embodiment verifies the online adjustment capability of the method of the present invention in dynamic scenarios.

[0070] During the control cycle Within the process, the controller first removes completed tasks, retaining the process paths of tasks still being processed in the previous cycle. For newly arriving tasks, the controller calculates the normalized bottleneck load for each candidate process path. The candidate process path with the smallest normalized bottleneck load is selected as the initial allocation result.

[0071] If a fairness violation occurs after a new job is added, the controller attempts to migrate the continuous job within a preset reconfiguration budget of B=2 or B=5. After each migration, the fair tick capability is recalculated, and a decision is made on whether to retain the migration action based on the tiered objectives (first reduce the number of violating jobs, then increase the minimum fairness retention ratio, and then increase the low quantile fairness retention ratio).

[0072] The initial allocation of new job tasks is not counted in the reconfiguration count of adjacent cycles to reduce the impact of frequent order insertions on the stability of existing job tasks. During dynamic order insertions or equipment degradation, the controller inherits the execution status of continuous job tasks, calculates the bottleneck load of candidate process paths, generates within-budget adjustment actions, issues and executes control commands, and provides feedback to verify cycle time, buffering, or alarms. If the fairness threshold is not met and the budget is not exhausted, local repair continues; otherwise, a feasibility flag and violation ratio are output. When there are no improvement actions, the budget is exhausted, there are insufficient candidate process paths, or there are insufficient equipment control entries within the control cycle, the controller maintains the process path control of the previous version and outputs the status, including unrepaired violation tasks and budget consumption.

[0073] This embodiment verifies the stability of the method of the present invention in a dynamic production line. By limiting the reconfiguration budget for each cycle, it effectively avoids frequent changes in control commands and state transitions caused by pursuing instantaneous optimality, thereby reducing the operational risks brought about by changes in industrial site rules.

[0074] Example 5: Coordinated Deployment of General Control or Regulation Systems and Industrial Field Control Systems This embodiment illustrates the actual deployment method of the system of the present invention; its system architecture is described in [reference needed]. Figure 1 .

[0075] The system of this invention can be deployed in a centralized production line controller, manufacturing execution system, robot cell controller, or edge process path control service. The controller, as a control functional unit of a general control or regulation system, operates a reference fair cycle time capability calculator, a risk control quantity calculator, a candidate process path manager, a regulation action generator, a fairness evaluator, and a control command issuer.

[0076] Industrial robots, programmable logic controllers (PLCs), or robot unit controllers execute control commands as controlled execution units. Low-level motion control, safety interlocks, equipment protection logic, and fieldbus status feedback continue to be handled by the original controllers of each device, completing motion interlocking, status feedback, and dynamic adjustment of cycle time within a given process path execution state.

[0077] The controller collects the status of the controlled industrial equipment through industrial field management interfaces, robot controllers, programmable logic controllers, conveyor line controllers, buffer station sensors, fixture status sensors, equipment utilization statistics or manufacturing execution systems, and generates and issues control commands according to the control cycle.

[0078] Candidate process paths can be fixed K process paths, dynamically generated sets of shortest process paths, segmented process paths, or sets of next workstations directly returned by industrial equipment.

[0079] The fair cycle time capability solver can employ incremental water level filling, hierarchical linear programming, bottleneck station elimination, distributed approximate iteration, or edge controller acceleration methods, depending on the production line scale and real-time requirements.

[0080] Fairness threshold It can be dynamically set according to order batch, synchronized order batch, order type, priority or control period, or segmented thresholds can be used.

[0081] Online budget B can be measured by the number of migration tasks, the number of control table item changes, the number of workstation switches, the proportion of process path execution status changes, or the amount of control instruction updates.

[0082] This embodiment demonstrates that the system of the present invention can work well with existing industrial field control systems, achieving upper-level closed-loop control without changing the underlying device interface protocol, exhibiting good compatibility and scalability. The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for fair cycle time control in a robot collaborative production line, applied to a general control or regulation system, wherein the general control or regulation system includes a controller, a status monitoring unit, a control command issuing unit, and a feedback verification unit, characterized in that, Includes the following steps: Step S1: The status monitoring unit monitors or tests the industrial automation production line equipment group and collects information such as workstation topology, equipment or workstation capacity, workstation status, buffer status, robot posture, fixture status, work-in-process position, and task status. Step S2: Generate a set of candidate process path execution states that can be executed by the controlled industrial equipment for the job task; Step S3: Establish a reference control model to characterize inlet-side constraints, outlet-side constraints and key process constraints, and calculate the reference fair cycle time capability of each task based on the reference control model. The reference fair cycle time capability is the cycle time capability of each task obtained by max-min fair calculation under the reference control model. Step S4: Based on the single-process path execution status under the current control strategy and the equipment or workstation capacity constraints, calculate the actual fair cycle time capability of each task. The actual fair cycle time capability is the cycle time capability of each task obtained by maximum and minimum fair calculation under the single-process path execution status under the current control strategy and the equipment or workstation capacity constraints. Step S5: Based on the deviation between the actual fair beat rate capability and the reference fair beat rate capability, determine the fairness maintenance ratio, the set of non-compliant tasks, and the synchronization batch tail control index to form a feedback deviation amount; the fairness maintenance ratio is the ratio of the actual fair beat rate capability to the reference fair beat rate capability. When the fairness maintenance ratio is lower than the preset fairness threshold, the corresponding work task is identified as a fairness violation task, and the set of violation tasks is composed of the fairness violation tasks; wherein, the synchronous batch tail control index is determined by the maximum value of the ratio of the remaining processing volume of each work task in the synchronous order batch to the actual fair cycle capacity, and is used to characterize the tail waiting risk of the controlled production batch. Step S6: Calculate the risk control quantity of the controlled task based on at least two of the following: entry station competition degree, exit station competition degree, candidate process path overlap degree, normalized bottleneck load, and order synchronization importance. The entry station competition degree represents the degree of competition for the task at the entry station; the exit station competition degree represents the degree of competition for the task at the exit station; the candidate process path overlap degree represents the degree to which candidate process paths share intermediate equipment or stations; the order synchronization importance represents the degree to which the task belongs to the synchronous order batch's synchronous completion requirement, or the degree of influence of the task on the completion time of the synchronous order batch; the normalized bottleneck load is determined by the maximum value among the ratios of the sum of the reference load of each equipment or station in the candidate process path and the reference fair takt capacity of the task to be assigned, and the capacity of the corresponding equipment or station. The reference load is the sum of the reference fair takt capacities of all tasks currently using that equipment or station. Step S7: Based on the risk control amount and according to the hierarchical feedback adjustment target, generate adjustment actions. The adjustment actions include process path execution status selection, robot task reallocation, buffer release, conveyor line diversion, workpiece reordering, or bottleneck station rerouting. The hierarchical feedback adjustment target refers to the target of selecting local adjustment actions according to the priority order of the number of non-compliant tasks, minimum fairness maintenance ratio, low percentile fairness maintenance ratio, and synchronous batch tail control index. The minimum fairness maintenance ratio is the minimum value among all work task fairness maintenance ratios; the low percentile fairness maintenance ratio is the low percentile value among all work task fairness maintenance ratios. Step S8: Within the preset repair budget or reconfiguration budget, the control command issuing unit issues robot operation program selection instructions, programmable logic controller variables, conveyor line diversion instructions, buffer release instructions, fixture switching instructions, or automatic guided transport equipment scheduling instructions to the controlled industrial equipment. Step S9: The feedback verification unit updates the control strategy based on the running status feedback after the instruction is issued.

2. The method according to claim 1, characterized in that, The reference control model is the reference input generation model of the general control or regulation system, including a virtual balanced production line model. The virtual balanced production line model retains the production line inlet capacity, production line outlet capacity, key process sequence and safety interlock constraints, and ignores temporary shared conflicts of non-critical intermediate stations. The virtual balanced production line model is used to generate the reference fair cycle time capability of each work task.

3. The method according to claim 1, characterized in that, The risk control parameters are used to determine the processing order of tasks and to construct a conflict diagram between tasks.

4. The method according to claim 1, characterized in that, The step S7, which generates the process path execution status based on the risk control quantity and selects the adjustment action, specifically includes: Work tasks are processed in descending order of risk control level; For each pending job task, calculate the conflict load, normalized bottleneck load, process path length, control rule change cost, and synchronous order batch impact of each candidate process path, and then weight them to obtain a comprehensive score; the synchronous order batch impact represents the influence of the candidate process path on the tail control index of the synchronous order batch. The overall score is obtained by weighting the conflict load, normalized bottleneck load, process path length, control rule change cost, and the impact of synchronous order batches; wherein, the greater the impact of conflict load, normalized bottleneck load, process path length, control rule change cost, and synchronous order batches, the higher the overall score; The execution state of the candidate process path with the lowest comprehensive score is determined, and the execution state of the candidate process path is converted into the execution control command of the controlled industrial equipment.

5. The method according to claim 1, characterized in that, In step S7, the set of work tasks is divided into a protection task set and a core task set. The protection task set consists of work tasks for which the fairness maintenance ratio needs to be prioritized to be no less than a preset fairness threshold. The core task set consists of work tasks for which the total production line capacity needs to be prioritized to be increased and the bottleneck load needs to be reduced. The protection task set generates adjustment actions based on the fairness takt maintenance target, and the core task set generates adjustment actions based on the capacity improvement target. The fairness takt maintenance target refers to the goal of ensuring that the fairness maintenance ratio of work tasks in the protection task set is no less than the preset fairness threshold. The capacity improvement target refers to the goal of increasing the total production line capacity corresponding to the core task set and reducing the bottleneck load. When there are any fairness violation tasks, local adjustment actions are selected according to the hierarchical feedback adjustment target.

6. The method according to claim 1, characterized in that, Within a dynamic control cycle, the path execution state of the continuous task from the previous control cycle is inherited; the continuous task is a task that existed in the previous control cycle and is still being processed in the current control cycle; for newly arriving tasks, the newly arrived task is a task newly added in the current control cycle; the normalized bottleneck load of each candidate process path is calculated, and the path execution state with the minimum normalized bottleneck load is determined; when a fair violation task is detected, only the continuous task is subject to within-budget control migration, and the number of migrations does not exceed the preset reconfiguration budget; The normalized bottleneck load is determined by the maximum value among the ratios of the sum of the reference load of each device or workstation in the candidate process path and the reference fair takt capacity of the task to be assigned to the capacity of the corresponding device or workstation; wherein, the reference load is the sum of the reference fair takt capacity of each task currently using the device or workstation, and the capacity is the capacity value in the device or workstation capacity constraint.

7. A general control or adjustment system for fair cycle control in a robot collaborative production line, characterized in that, It includes a controller, a status monitoring unit, a reference control model unit, a deviation calculation unit, a risk control quantity calculation unit, a regulation action generation unit, a budget constraint execution unit, a control command issuance unit, and a feedback verification unit; The status monitoring unit is used to monitor or test the equipment group of industrial automated production line, and collect the workstation topology, equipment or workstation capacity, workstation status, buffer status, robot posture, fixture status, work-in-process position and task status. The controller is used to generate a set of candidate process path execution states that can be executed by the controlled industrial equipment for the job task; The reference control model unit is used to establish a reference control model to characterize inlet-side constraints, outlet-side constraints and key process constraints, and to calculate the reference fair cycle time capability of each task. The reference fair takt capability is the takt capability of each task obtained by max-min fairness calculation under the reference control model; The deviation calculation unit is used to calculate the actual fair cycle time capability based on the single-process path execution status and equipment or workstation capacity constraints under the current control strategy. The actual fair cycle time capability is the cycle time capability of each task obtained by maximum and minimum fairness calculation under the single-process path execution status and equipment or workstation capacity constraints under the current control strategy. Based on the deviation between the actual fair cycle time capability and the reference fair cycle time capability, the unit determines the fairness maintenance ratio, the set of non-compliant tasks, and the synchronization batch tail control index to form a feedback deviation. The fairness maintenance ratio is the ratio of the actual fair cycle time capability to the reference fair cycle time capability. When the fairness maintenance ratio is lower than the preset fairness threshold, the corresponding job task is identified as a fairness violation task, and the set of violation tasks is composed of the fairness violation tasks; the synchronous batch tail control index is determined by the maximum value of the ratio of the remaining processing volume of each job task in the synchronous order batch to the actual fair cycle capacity. The risk control quantity calculation unit is used to calculate the risk control quantity of the controlled task based on at least two of the following: entry station competition degree, exit station competition degree, candidate process path overlap degree, normalized bottleneck load, and order synchronization importance. The entry station competition degree represents the degree to which a task competes with other tasks for the same entry station. The export workstation competition degree indicates the extent to which a task competes with other tasks for the same export workstation; the candidate process path overlap degree indicates the extent to which candidate process paths among tasks share intermediate equipment or workstations; the order synchronization importance indicates the degree to which the task belongs to the synchronous order batch's synchronous completion requirement, or the degree of influence of the task on the completion time of the synchronous order batch; the normalized bottleneck load is determined according to the maximum value of the ratio of the sum of the reference load of each equipment or workstation in the candidate process path and the reference fair takt capacity of the task to be assigned to the capacity of the corresponding equipment or workstation; wherein, the reference load is the sum of the reference fair takt capacity of each task currently using the equipment or workstation; The adjustment action generation unit is used to generate adjustment actions based on the risk control amount and according to the hierarchical feedback adjustment target. The adjustment actions include process path execution state selection, robot task reallocation, buffer release, conveyor line diversion, workpiece reordering, or bottleneck station clearance. The hierarchical feedback adjustment target refers to the target of selecting local adjustment actions according to the priority order of the number of non-compliant tasks, minimum fairness maintenance ratio, low percentile fairness maintenance ratio, and synchronous batch tail control index. The minimum fairness maintenance ratio is the minimum value among all work task fairness maintenance ratios. The low percentile fairness maintenance ratio is the low percentile value among all work task fairness maintenance ratios. The budget constraint execution unit is used to perform control migration within a preset repair budget or reconfiguration budget; The control command issuing unit is used to issue robot operation program selection instructions, programmable logic controller variables, conveyor line diversion instructions, buffer release instructions, fixture switching instructions or automatic guided transportation equipment scheduling instructions to the controlled industrial equipment. The feedback verification unit is used to update the control strategy based on the feedback of the operating status.

8. An electronic device comprising a processor and a memory, wherein the memory stores a computer program that, when executed on the processor, causes the electronic device to perform the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, the computer program, when executed on a processor, causing the processor to perform the method according to any one of claims 1 to 6.

Images