An assembly workpiece production line automation management system and method

Abstract

This invention discloses an automated management system and method for assembly workpiece production lines, relating to the field of cloud-edge collaborative technology. The invention integrates assembly operation observation data and cloud-edge scheduling status data, using edge compensation trigger points that have completed time-base correction as event anchor points. It constructs compensation occlusion event fragments according to assembly action affiliation and edge state version. Based on these fragments, it generates compensation occlusion results and degradation overflow fragments, and uses the degradation overflow fragments as propagation starting points to construct a compensation constraint beat propagation relationship graph, generating degradation anomaly source labels. Furthermore, it combines edge state version with cloud-edge state effective interval verification, outputting workstation scheduling restriction results, pending review scheduling instructions, scheduling cancellation results, and anomaly rollback benefit results. This system can identify hidden degradation states masked by edge compensation and reduce cloud-based scheduling control mismatch.

Description

Technical Field

[0001] This invention relates to the field of cloud-edge collaboration technology, specifically to an automated management system and method for assembly workpiece production lines. Background Technology

[0002] Assembly lines typically include continuous workstations such as press-fitting, tightening, dispensing, vision inspection, robot handling, fixture positioning, buffer transfer, and workpiece re-inspection. On-site edge controllers control assembly actions in real time based on pressure, displacement, torque, visual deviation, cycle time, and inspection results. When deviations occur, they provide local compensation for press-fitting stroke, tightening torque, dispensing time, robot path, or visual positioning reference. Cloud platforms typically access data from MES, SCADA, APS, and equipment maintenance systems to summarize work order progress, output, pass rate, alarm records, workstation cycle time, and maintenance status. Based on this, they perform capacity analysis, anomaly statistics, maintenance scheduling, and production task adjustments. In multi-model workpiece mixed-line production scenarios, different workpieces have different assembly cycle times, process paths, inspection standards, and compensation boundaries. Workstations are continuously coupled through buffer zones, transfer mechanisms, and AGVs. Changes in the status of any workstation can affect upstream and downstream workstations through waiting time, buffer occupancy, and flow cycle time.

[0003] Existing assembly line management technologies typically employ PLCs or edge controllers for partial closed-loop control, SCADA to collect equipment status and alarm information, MES to record process completion and production results, and APS or cloud management platforms for scheduling and adjustments based on orders, capacity, and delivery dates. Predictive maintenance solutions often rely on vibration, current, temperature, alarm frequency, downtime records, or quality defect rates to assess equipment risk. While these solutions can record and analyze explicit faults, abnormal results, and changes in production schedule, edge-side compensation actions in existing cloud management data are usually stored as control execution records and are rarely used for further table processing. During the equipment degradation process, in actual production, fixture wear, actuator drift, positioning reference offset, or changes in pressing resistance may initially be absorbed by edge compensation actions, keeping the output, pass rate, and alarm count seen in the cloud within the normal range. However, as the compensation margin continues to be consumed, anomalies further manifest as workstation cycle time drift, increased re-inspection, changes in buffer occupancy, and changes in upstream and downstream waiting times. Without constraints on the consumption of compensation margin, such cycle time changes may be attributed to reasons such as incoming material fluctuations, downstream congestion, common bottlenecks, or data upload delays. Consequently, under certain operating conditions, this can lead to discrepancies between maintenance triggering, scheduling restrictions, and anomaly cancellation control and the real-time status on the edge side. Summary of the Invention

[0004] To achieve the above objectives, the present invention provides the following technical solution: An automated management system for an assembly line includes: The compensation occlusion segment construction module connects assembly operation observation data and cloud-edge scheduling status data. Taking the edge compensation trigger point that completes time base correction as the event anchor point, it locks the execution status before and after compensation according to the assembly action affiliation and edge status version. It constructs a compensation occlusion closure relationship that represents the execution deviation being absorbed by edge compensation and forming the beat response in the quality release state, and generates compensation occlusion event segments. The degradation overflow identification module performs deviation in the same direction conversion, compensation occupancy characterization, and quality release comparison with stable beat for the compensation occupancy event segment, generates compensation occupancy results, and generates degradation overflow segments when the compensation occupancy results are in an upward state within the sliding production window and trigger beat overflow state. The propagation attribution module takes the degraded spillover fragment as the propagation starting point, determines the affected fragments according to the validity of the production line flow sequence and the credibility of the cloud edge state, and performs compensation constraint attribution on the propagation relationship between the degraded spillover fragment and the affected fragments, forming a compensation constraint beat propagation relationship diagram, and generating degraded anomaly source labels. The cloud-edge control verification module verifies the effective range of the execution state of cloud scheduling instructions in the pre-execution state based on the degradation anomaly source label, the compensation constraint beat propagation relationship diagram, and the edge state version. It generates scheduling verification results, which include at least one of the workstation scheduling restriction results and scheduling instructions to be reviewed. When the recovery closure result formed by the recovery observation is valid, it generates scheduling cancellation results and anomaly rollback benefit results.

[0005] Furthermore, using the edge compensation trigger point that has completed time base correction as the event anchor point, the process includes: performing time base correction on assembly operation observation data and cloud edge scheduling status data based on at least one of edge node clock synchronization records, cloud-edge communication logs, and a unified clock source for the production line; using the edge compensation trigger point after time base correction as the segment start boundary, the execution closing time of the same assembly action as the segment first closing boundary, and the quality inspection result return time as the segment second closing boundary; performing consistency matching on the segment start boundary, segment first closing boundary, and segment second closing boundary according to the assembly action attribution and edge state version to form segment boundaries; when the quality inspection result is not returned, the corresponding segment is marked as an unclosed segment, and delayed closure is performed according to the assembly action attribution and edge state version after the quality inspection result is returned.

[0006] Furthermore, the process of forming the compensation masking closure relationship of the cycle response under the quality release state includes: within the segment boundary, performing a dimensionless conversion based on the corresponding execution reference and deviation boundary on the execution feedback quantity before compensation triggering to obtain the execution deviation before compensation; performing an occupancy mapping based on the corresponding calibration compensation upper limit on the edge control compensation record to obtain the compensation action description quantity; when the quality state after compensation meets the release conditions, mapping the cycle result after compensation to the cycle response under the quality release state, and closing the execution deviation before compensation, the compensation action description quantity, and the cycle response under the quality release state according to the segment boundary to generate a compensation masking event segment used to characterize the execution deviation being absorbed by edge compensation; wherein, the execution reference, deviation boundary, and calibration compensation upper limit are derived from equipment calibration records, process specifications, commissioning and acceptance records, or historical stable production samples.

[0007] Furthermore, the compensation occlusion fragment construction module is also used to configure fragment status markers for compensation occlusion event fragments. Fragment status markers include closed fragment markers, unclosed fragment markers, version to be verified markers, and compensation boundary to be verified markers. When the edge state version referenced by the cloud scheduling instruction is inconsistent with the edge state version when the assembly action is actually executed, a version to be verified marker is configured. The version to be verified marker is used to restrict the direct execution of the cloud scheduling instruction. When the compensation parameter does not have a calibrated compensation upper limit or the calibrated compensation upper limit is zero, a compensation boundary to be verified marker is configured. Compensation occlusion event fragments with compensation boundary to be verified markers do not participate in the subsequent degradation overflow identification process based on the calibrated compensation upper limit to form a compensation approximation result. The execution deviation before compensation, the beat response under the quality release state, and the flow state within the fragment boundary are retained as the basis for judging output occlusion, overflow accompanying changes, and boundary verification.

[0008] Furthermore, the operation of generating compensation masking results includes: establishing a set of similar stable benchmark samples according to the assembly action affiliation; forming a stable operation reference boundary from the set of similar stable benchmark samples for compensation masking identification and degradation spillover identification; forming compensation proximity results based on the compensation action description quantity; forming compensation dependency results based on whether the compensation action occurs within the sliding production window, the continuous compensation state, and the compensation duration state; forming output masking results based on segments where the execution deviation before compensation exceeds the stable operation reference boundary and the quality state after compensation meets the release conditions; and merging the compensation proximity results, compensation dependency results, and output masking results in the same direction to generate compensation masking results. Among these, the set of similar stable benchmark samples is formed from sample records confirmed by production stability. The sample records include equipment commissioning and acceptance records, continuous stable production cycle records, stable operation samples confirmed by process procedures, verification passed samples confirmed by quality verification procedures, and verification passed samples confirmed by maintenance acceptance records.

[0009] Furthermore, the operation of generating degraded spillover segments includes: when the compensation masking result meets the compensation masking establishment condition formed by similar stable benchmark samples, and the compensation masking result meets the upward trend condition formed by the comparison of the front and rear segments of the sliding production window, the compensated cycle time result is compared with the stable operation reference boundary, and the flow-related changes that deviate from the stable operation reference boundary within the segment boundary are merged into spillover-related changes; a cycle time spillover state is formed based on the compensated cycle time offset and spillover-related changes; when the cycle time spillover state meets the spillover establishment condition formed by at least one of the stable production samples, process cycle time table, quality review procedure, and historical anomaly review samples, a degraded spillover segment is generated; segments whose quality status after compensation does not meet the release conditions do not participate in the formation of the output masking result, but are retained as compensation masking failure segments, which are used for maintenance confirmation and recovery observation processes.

[0010] Furthermore, the operation of compensating for the attribution of the propagation relationship between the degraded spillover fragment and the affected fragment includes: taking the degraded spillover fragment as the propagation starting point, determining the propagation observation range based on the production line flow path, and identifying the affected fragment within the propagation observation range; performing flow timing verification, propagation direction verification, compensation spillover correlation verification, and cloud-edge timing reliability verification on the propagation relationship between the degraded spillover fragment and the affected fragment; flow timing verification is used to confirm that the affected fragment is within the propagation observation range, propagation direction verification is used to confirm that the direction of state change matches the production line flow path, and compensation spillover correlation verification is used to confirm that the change of the affected fragment is caused by the compensation occlusion result in the degraded spillover fragment and the beat spillover state. Support, cloud-edge time sequence reliability verification is used to exclude propagation relationships formed solely by cloud-received time sequences; when a propagation relationship passes verification and is not fully explained by non-degenerate anomaly source labels, a compensation constraint propagation edge is established, and a compensation constraint beat propagation relationship graph and degenerate anomaly source labels are formed based on the compensation constraint propagation edge. Non-degenerate anomaly source labels are used to characterize the explanatory relationship of non-degenerate causes on beat changes; wherein, the propagation observation range is formed by standard flow time, buffer release time and maximum allowable waiting time of production line, and the standard flow time, buffer release time and maximum allowable waiting time of production line are derived from at least one of the following: process beat table, buffer design record, transfer mechanism record and historical stable production sample.

[0011] Furthermore, the operation of verifying the basis for the formation of degenerate anomaly source tags based on non-degenerate anomaly source tags includes: performing a comprehensive interpretation verification on the non-degenerate anomaly source tags; when the non-degenerate anomaly source tags can cover all anomaly changes of the same event anchor point, the same state segment, and the same propagation edge, no degenerate anomaly source tag is generated; when the non-degenerate anomaly source tags only interpret part of the propagation edge, only the interpreted propagation edge is excluded, and the uninterpreted compensation constraint propagation edge is retained; for the same candidate propagation relationship, only one exclusion rule is used as the main exclusion basis, and the other exclusion rules are used as auxiliary explanations, without repeatedly weakening the basis for the formation of degenerate anomaly source tags; wherein, the non-degenerate anomaly source tags include tags used to characterize at least one of the following interpretation relationships of workpiece model switching, incoming material fluctuation, downstream congestion, buffer occupation, and data upload latency on cycle time changes.

[0012] Furthermore, the operations for generating recovery closure results, scheduling cancellation results, and anomaly rollback benefit results include: generating edge anomaly state records based on the degenerate anomaly source label and the compensation constraint beat propagation relationship diagram; determining the state influence range based on the compensation constraint propagation edge; forming the edge state effective interval based on the degenerate spillover fragment generation time, the degenerate anomaly source label generation time, and the state natural failure time; updating the end boundary of the edge state effective interval with the recovery closure result generation time after the recovery closure result is generated, wherein the state natural failure time is formed by the propagation observation range, the maximum allowable waiting time of the production line, and the beat response boundary in the stable operation reference boundary; verifying the edge state version, planned execution time, and effective workstation range referenced by the cloud scheduling instruction in the pre-execution state, generating scheduling verification results, which include workstation scheduling restriction results and pending... At least one of the review scheduling instructions is used to configure processing action markers for the compensation occlusion event segments in subsequent recovery observations. The processing action markers are used to distinguish between endogenous state changes and external disturbances caused by at least one of the workstation scheduling restrictions, maintenance confirmations, and pending review processing. During the recovery observation process, the compensation occlusion state, the cycle overflow state, and the flow-accompanying state all revert to the recovery benchmark formed by similar stable benchmark samples, maintenance acceptance records, or quality review procedures, satisfying the continuous quality release condition formed by at least one of the quality review procedures and maintenance acceptance records. At the same time, the recovery state version synchronization is completed as the recovery closure verification condition. When the recovery closure verification passes, a recovery closure result is generated. Based on the recovery closure result, a scheduling cancellation result and anomaly rollback benefit result are generated. The anomaly rollback benefit result is bound to the degradation anomaly source label, state influence range, compensation constraint propagation edge, and scheduling cancellation result.

[0013] An automated management method for an assembly line includes the following steps: By accessing assembly operation observation data and cloud-edge scheduling status data, and taking the edge compensation trigger point that completes time base correction as the event anchor point, the execution status before and after compensation is locked according to the assembly action affiliation and edge status version. A compensation occlusion closure relationship is constructed to represent the execution deviation being absorbed by edge compensation and forming the beat response in the quality release state, and compensation occlusion event fragments are generated. Perform deviation in the same direction conversion, compensation occupancy characterization, and quality release comparison with stable beat on the compensation occupancy event segment to generate compensation occupancy results. When the compensation occupancy result is in an upward state within the sliding production window and triggers beat overflow state, generate degenerate overflow segment. Starting from the degraded spillover fragments, the affected fragments are determined according to the validity of the production line flow sequence and the credibility of the cloud edge status. The propagation relationship between the degraded spillover fragments and the affected fragments is attributed with compensation constraints, forming a compensation constraint beat propagation relationship diagram, and generating a degraded anomaly source label. Based on the degenerate anomaly source label, the compensation constraint beat propagation relationship diagram, and the edge state version, the effective interval of the execution state of the cloud scheduling instruction in the pre-execution state is verified, and a scheduling verification result is generated. The scheduling verification result includes at least one of the workstation scheduling restriction result and the scheduling instruction to be reviewed. When the recovery closure result formed by the recovery observation is valid, a scheduling cancellation result and anomaly rollback benefit result are generated.

[0014] This invention provides an automated management system and method for assembly workpiece production lines, which has the following beneficial effects: 1. This invention solves the problem of edge compensation actions masking gradual degradation of equipment while maintaining the quality release state by using the edge compensation trigger point that completes time base correction as the event anchor point and constructing compensation occlusion event segments according to the assembly action affiliation and edge state version. As a result, the execution deviation before compensation, the description quantity of the compensation action, and the cycle response in the quality release state can be limited to the same segment boundary, realizing the traceability of the compensation occupation process behind the qualified output, and also providing a feasible state basis for subsequent identification of hidden degradation.

[0015] 2. Based on the already formed compensation occlusion event fragments, this invention further solves the problem of difficulty in distinguishing between single compensation fluctuations, ordinary beat fluctuations and continuous compensation dependencies by linking the compensation occlusion results, the upward trend within the sliding production window, the beat offset after compensation and the accompanying changes in overflow. This processing extends the degradation identification from the compensation absorption state within a single workstation to the upstream and downstream beat propagation state, realizing the accurate generation of degradation overflow fragments, and also realizing the continuous evidence closure between compensation dependencies, beat drift and accompanying changes in flow.

[0016] 3. Based on the aforementioned degenerate spillover fragments, this invention constructs a compensating constraint beat propagation relationship graph and performs cloud-edge control verification by combining degenerate anomaly source labels, edge state versions, and edge state effective intervals. This solves the problem of cloud-based scheduling restrictions based on apparent beat anomalies or expired edge states. This process binds anomaly identification, propagation attribution, scheduling restrictions, pending review processing, recovery closure, and anomaly rollback benefits into the same control chain. It realizes the version constraint output of workstation scheduling restrictions and pending review scheduling instructions, and also realizes the attribution of restored scheduling cancellation and anomaly rollback benefits. Attached Figure Description

[0017] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a system framework diagram of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1: Please see Figure 1 This embodiment provides an automated management system for an assembly workpiece production line, including: a compensation occlusion segment construction module, a degradation overflow identification module, a propagation attribution module, and a cloud-edge control verification module.

[0020] In this embodiment, the automated management system for the assembly production line accesses assembly operation observation data and cloud-edge scheduling status data through a compensation occlusion segment construction module. The assembly operation observation data serves as the input for the formation of compensation occlusion event segments, and is used to characterize the execution status, quality release status, cycle time response, and flow status within the segment boundary before and after edge compensation is triggered. The cloud-edge scheduling status data serves as the version reference input, and is used to characterize the reference relationship between cloud scheduling instructions and edge status versions. The above data is provided by the production line edge controller, workstation control equipment, quality inspection equipment, buffer detection unit, transfer equipment, cloud scheduling platform, and edge gateway, and forms time-stamped observation records.

[0021] The compensation occlusion segment construction module performs time base correction on assembly operation observation data and cloud-edge scheduling status data based on at least one of edge node clock synchronization records, cloud-edge communication logs, and a unified clock source for the production line. This yields edge compensation trigger points, assembly action execution closing times, quality inspection result return times, and cloud scheduling instruction reference times all under the same time base. Time base correction is used to avoid segment boundary mismatches caused by deviations between edge-side acquisition times, cloud-received times, and cloud scheduling instruction times. The residual time deviation after time base correction is determined based on edge node clock synchronization accuracy, the shortest event interval for the same workstation, equipment operating procedures, and... The residual time deviation is determined by the cloud-edge communication log, preferably between 1ms and 100ms, and not greater than one-tenth of the shortest event interval for the same workstation. For robot positioning, vision guidance, or high-speed grasping workstations, the residual time deviation is preferably between 1ms and 20ms. For pressing, tightening, dispensing, or pressure holding workstations, the residual time deviation is preferably between 20ms and 100ms. When the number of historical communication logs is insufficient, the time base correction target is determined by the edge gateway acceptance record, the accuracy of the unified clock source on the production line, and the equipment operation procedures. When the residual time deviation after time base correction does not meet the segment boundary recognition requirements, the corresponding segment does not directly enter the subsequent degradation overflow recognition.

[0022] For segments that do not directly enter the degradation overflow identification process due to residual time deviation failing to meet the segment boundary identification requirements, the compensation occlusion segment construction module configures them as time base verification segments and retains their assembly action attribution, edge state version, original generation time on the edge side, cloud reception time, compensation trigger record, and quality inspection return record. When subsequent time base verification is completed through edge node clock synchronization records, cloud-edge communication logs, or the unified clock source of the production line, if the segment can match the same assembly action attribution and the same edge state version, it is updated as a closed segment or delayed closed segment and continues to enter the compensation occlusion identification or recovery observation process. This processing is used to avoid erroneously excluding real compensation occlusion states or abnormal rollback states as invalid data due to time base anomalies.

[0023] After completing the time base correction, the compensation occlusion segment construction module uses the edge compensation trigger point as the starting anchor point of the compensation occlusion event segment. The edge compensation trigger point is the moment when the edge controller triggers compensation control based on the current execution feedback state. The compensation control includes at least one of press stroke compensation, tightening torque compensation, dispensing time compensation, robot path compensation, and visual positioning correction. Using the edge compensation trigger point as the event anchor point is used to limit the execution deviation occurrence, compensation action intervention, post-compensation quality state, cycle response in the quality release state, and cloud scheduling version reference within the same assembly action boundary, so that the edge compensation occupancy process behind the quality release result can be identified later.

[0024] The compensation occlusion segment construction module performs consistency matching on the segment start boundary, segment first closed boundary, and segment second closed boundary according to the assembly action attribution and edge state version to form segment boundaries. Among them, the segment start boundary is the edge compensation trigger point after the completion of time base correction; the segment first closed boundary is the execution closing time of the same assembly action, which includes at least one of the following: press-fit completion time, tightening completion time, dispensing completion time, robot handling arrival time, and visual positioning correction completion time; the segment second closed boundary is the quality inspection result return time. The assembly action attribution is determined based on the workpiece number, workpiece model, station number, and process action identifier. The edge state version is determined based on the edge node number, station number, process action identifier, status update time, and version number. When the segment start boundary, segment first closed boundary, and segment second closed boundary all satisfy the same assembly action attribution and the same edge state version, the corresponding segment is entered as a closed segment for subsequent compensation occlusion recognition.

[0025] When the quality inspection result is not returned, the compensation occlusion segment construction module marks the corresponding segment as an unclosed segment. Unclosed segments do not directly participate in the formation of the compensation occlusion result. When the quality inspection result is returned, the compensation occlusion segment construction module performs delayed closure according to the assembly action attribution and edge state version. If the returned quality inspection result is consistent with the assembly action attribution and edge state version corresponding to the segment's starting boundary, the unclosed segment is updated to a closed segment. If the returned quality inspection result cannot match the assembly action attribution or edge state version, the unclosed segment mark is retained, and it does not directly enter the subsequent degradation overflow identification. The delayed closure waiting time is determined based on the response time of the quality inspection equipment, the cycle time of the subsequent inspection station, and the maximum dwell time of the production line buffer area. Preferably, the time range corresponding to the 95th to 99th percentile of the historical quality inspection return time of the same station is used. When the number of historical samples is insufficient, at least one of the quality inspection equipment acceptance records, process specifications, and production line buffer area design records is used to determine the delayed closure waiting time.

[0026] Within the segment boundary, the compensation masking segment construction module performs a dimensionless conversion on the execution feedback quantity before compensation triggering, based on the corresponding execution reference and deviation boundary, to obtain the execution deviation before compensation. The execution feedback quantity includes at least one of pressure, displacement, torque, dispensing amount, robot positioning deviation, vision positioning deviation, and holding pressure time. Since different execution feedback quantities have different physical units, the compensation masking segment construction module first calculates the deviation of the execution feedback quantity relative to the execution reference based on the execution reference of the corresponding workstation, corresponding process action, and corresponding workpiece model, and then converts the individual deviation results into dimensionless values ​​with the same direction based on the deviation boundary. As a result, the single deviation result is preferably mapped to the range of 0 to 1. The larger the value, the higher the degree of deviation of the execution feedback quantity from the execution reference. When the execution feedback quantity exceeds the deviation boundary, it is truncated according to the upper limit value corresponding to the deviation boundary. The execution reference comes from at least one of the equipment calibration record, process specification, machine commissioning and acceptance record and historical stable production sample. The deviation boundary comes from at least one of the quality acceptance tolerance, equipment calibration boundary, machine commissioning and acceptance record and historical stable production sample allowable fluctuation boundary. The execution deviation before compensation is used as the input for subsequent output occlusion judgment to determine whether the execution deviation has been absorbed by the edge compensation action.

[0027] The compensation occupancy segment construction module performs occupancy mapping on the edge control compensation records based on the corresponding calibrated compensation upper limit to obtain the compensation action description quantity. Within the same segment boundary, the compensation occupancy segment construction module reads the control setpoint before compensation, the control setpoint after compensation, the compensation trigger time, the compensation execution duration, the compensation trigger condition, and the calibrated compensation upper limit. It maps the change in each type of compensation parameter to the calibrated compensation upper limit under the corresponding workstation, the corresponding process action, and the corresponding workpiece model to form a single compensation occupancy result. Then, according to the weight determined by the process action quality characteristics, control priority, or historical defect attribution records, the single compensation occupancy results are merged into a compensation action description quantity. The single compensation occupancy result is preferably mapped to the range of 0 to 1, with larger values ​​indicating higher values. The higher the degree of occupation of the calibration compensation upper limit by the corresponding compensation parameter, the higher the weight is. The weight is a non-negative weight, and the sum of the weights participating in the merging under the same process action is 1. The calibration compensation upper limit comes from at least one of the equipment calibration record, process procedure, commissioning and acceptance record and safety control boundary. The calibration compensation upper limit is greater than 0 and does not exceed the compensation range allowed by the equipment controller or the safety compensation boundary specified by the process procedure. When the compensation parameter does not have a calibration compensation upper limit or the calibration compensation upper limit is 0, the compensation occupancy segment construction module does not use the compensation parameter to form the compensation action description quantity, and configures the corresponding segment as a compensation boundary to be verified mark. The compensation action description quantity is used as the input for the subsequent compensation approximation result formation to characterize the degree to which the edge compensation capability is occupied.

[0028] The compensation occlusion segment construction module determines whether to form a clock response in the quality release state based on the post-compensation quality status. The post-compensation quality status is determined according to the process specifications, quality inspection standards, or equipment acceptance standards. When the post-compensation quality inspection result meets the release conditions of the corresponding process action, it is marked as a quality release state, and the post-compensation clock response is mapped to the clock response in the quality release state. When the post-compensation quality inspection result does not meet the release conditions of the corresponding process action, it is marked as not meeting the release conditions. The pre-compensation execution deviation, compensation action description, post-compensation clock response, and segment status of the segment are retained, but the post-compensation clock response is not mapped to the clock response in the quality release state. This segment is used for subsequent compensation occlusion failure segment identification, maintenance confirmation, and recovery observation processes.

[0029] The compensated beat result is determined based on the start trigger time and execution completion time of the same assembly action. It is used to represent the overall execution beat formed by the assembly action after edge compensation intervention. The compensated beat result is not used to characterize the execution time of the edge compensation action itself, but is used to further form the beat response in the quality release state when the quality state after compensation meets the release conditions, and serves as the input for the subsequent degradation overflow identification module to generate the compensated beat offset and degradation overflow segment.

[0030] Both the start trigger time and the execution completion time are derived from the assembly operation observation data after time base correction, and they should correspond to the same assembly action and the same edge state version. The start trigger time and the execution completion time use the same time unit, which is seconds or milliseconds. The compensated cycle time result is calculated according to the following formula: J_i=b_i-a_i; Where i is the assembly action index, used to distinguish different assembly actions, and does not represent the workstation number, workpiece number, or edge state version; J_i represents the compensated cycle time result of the i-th assembly action, which is a time quantity; a_i represents the start trigger time of the i-th assembly action, and b_i represents the execution completion time of the i-th assembly action. Both a_i and b_i are derived from the assembly operation observation data after time base correction and use the same time unit.

[0031] The formula determines the compensated cycle time result by the time difference between the execution completion time and the start trigger time. Its design logic is that the same assembly action may still form a quality release state after edge compensation intervention, but its overall execution cycle time may have shifted relative to the stable operation reference boundary. Therefore, it is necessary to calculate the compensated cycle time result with the start trigger time and execution completion time under the same time base so as to bind the quality release state and cycle time response to the same compensated occlusion event segment. The calculation result is used to support the subsequent mapping of the compensated cycle time result to the cycle time response in the quality release state, and further support the formation of the compensated occlusion result, compensated cycle time offset and degradation overflow segment in the degradation overflow identification module.

[0032] When the execution completion time is earlier than the start trigger time, or when the start trigger time and the execution completion time correspond to different assembly actions or different edge state versions, the compensation occlusion segment construction module does not calculate the compensation-after-beat result and configures the corresponding segment as an unclosed segment marker or a version pending verification marker. This processing is used to avoid distortion of the compensation-after-beat result due to time base abnormalities, segment mismatches, or version reference errors.

[0033] The compensation masking segment construction module closes the execution deviation before compensation, the description of the compensation action, and the cycle response in the quality release state according to the same segment boundary, generating a compensation masking event segment to characterize the execution deviation being absorbed by edge compensation. This compensation masking event segment is used to characterize that when there is already an execution deviation before compensation is triggered, the edge controller makes the assembly action still form a cycle response in the quality release state through compensation action. Compared with the production management method that only records production results, pass rate, or alarm count, the compensation masking event segment retains the compensation occupancy process behind the quality release state, providing a basis for subsequent identification of the gradual degradation of equipment being masked by edge compensation.

[0034] The compensation occlusion fragment construction module configures fragment status markers for compensation occlusion event fragments. These fragment status markers include closed fragment markers, unclosed fragment markers, version pending review markers, and compensation boundary pending review markers. A closed fragment marker is configured when the compensation occlusion event fragment has complete pre-compensation execution deviation, compensation action description, beat response under quality release status, and fragment boundaries. An unclosed fragment marker is configured when the quality inspection result is not returned. A version pending review marker is configured when the edge state version referenced by the cloud scheduling instruction is inconsistent with the edge state version actually executed during assembly. The version pending review flag is used to restrict the direct execution of cloud scheduling instructions. When the compensation parameter does not have a calibrated compensation upper limit or the calibrated compensation upper limit is 0, a compensation boundary pending review flag is configured. Compensation occlusion event segments with the compensation boundary pending review flag do not participate in the subsequent degradation overflow identification process based on the calibrated compensation upper limit to form a compensation approximation result. The execution deviation before compensation, the beat response under the quality release state, and the flow state within the segment boundary are retained as the basis for judging output occlusion, overflow accompanying changes, and boundary review, so as to avoid the overall exclusion of the actual execution deviation absorption process due to the lack of compensation upper limit.

[0035] Finally, the compensation occlusion fragment construction module encapsulates the fragment boundary, assembly action attribution, edge state version, pre-compensation execution deviation, compensation action description quantity, beat response under quality release state, flow state within the fragment boundary, version reference state of cloud scheduling instructions, and fragment state marker into compensation occlusion event fragments, forming a compensation occlusion event fragment set. The compensation occlusion event fragment set is used by the degradation overflow identification module to generate compensation occlusion results and degradation overflow fragments.

[0036] In this embodiment, the degradation overflow identification module receives the set of compensation occupancy event segments formed by the compensation occupancy segment construction module, performs deviation in the same direction conversion, compensation occupancy characterization, and quality release and stable beat comparison on the compensation occupancy event segments to generate compensation occupancy results; when the compensation occupancy result rises within the sliding production window and triggers a beat offset or overflow-related change after compensation, a degradation overflow segment is generated. This step is used to identify from the compensation occupancy event segments the state where the execution deviation is absorbed by edge compensation and the quality state after compensation still meets the release conditions, and to determine whether the state has overflowed to the beat response and flow-related changes.

[0037] The degradation overflow identification module first filters the set of compensated occlusion event segments into computable segments. Compensated occlusion event segments with closed segment markers and containing pre-compensation execution deviation, post-compensation quality status, and post-compensation beat results are included in the computable compensation occlusion event segment set. Compensated occlusion event segments with unclosed segment markers are included in the computable compensation occlusion event segment set after the quality detection result is returned and the delayed closure is completed. Compensated occlusion event segments with version pending verification markers retain their original execution facts on the edge side, but do not directly trigger cloud scheduling control. Compensated occlusion event segments with compensation boundary pending verification markers do not participate in the processing of compensation approximation results based on the calibrated compensation upper limit, but their pre-compensation execution deviation, beat response under quality release status, and flow status within the segment boundary are retained as the basis for judging the changes accompanying output occlusion and overflow. Through the above screening, the true compensation occlusion status is avoided from being filtered out due to quality detection delays, version inconsistencies, or missing compensation upper limits.

[0038] For computable compensation occlusion event fragments, the degradation overflow identification module establishes a set of similar stable benchmark samples according to the assembly action attribution. The assembly action attribution includes the correspondence between workpiece model, workstation number, process action identifier, and edge state version. The set of similar stable benchmark samples is formed by sample records confirmed by production stability. The sample records include equipment commissioning and acceptance records, continuous stable production cycle records, stable operation samples confirmed by process procedures, verification passed samples confirmed by quality verification procedures, and verification passed samples confirmed by maintenance acceptance records. The set of similar stable benchmark samples is used to form stable operation reference boundaries and serves as the sample basis for compensation occlusion establishment conditions, upward trend conditions, and overflow establishment conditions.

[0039] The number of samples in the same stable benchmark sample set is the number of valid assembly action samples, which is a dimensionless count. Preferably, the same stable benchmark sample set contains no less than 30 valid assembly action samples. For high-speed stations with a single-piece cycle time of less than 5 seconds, the same stable benchmark sample set contains no less than 100 valid assembly action samples. For small-batch or multi-model mixed-line production scenarios, the same stable benchmark sample set is formed using valid assembly action samples from the three most recent confirmed stable batches. The number of samples is determined based on the station cycle time, process statistics requirements, quality review procedures, and maintenance acceptance records. When the same stable benchmark sample set contains less than 100 valid assembly action samples, the stable operation reference boundary is not formed solely by percentile statistics results. Instead, the stable operation reference boundary is determined by combining the equipment commissioning and acceptance records, the stable operation samples confirmed by the process procedures, and the review samples confirmed by the maintenance acceptance records.

[0040] The degradation spillover identification module forms stable operating reference boundaries from a set of similar stable benchmark samples. These stable operating reference boundaries correspond to execution deviations, cycle time responses, and accompanying changes in the workflow. Stable operating reference boundaries for different categories are formed separately without direct merging across dimensions. For execution deviations before compensation, the stable operating reference boundaries are formed using the stable distribution of execution deviations before compensation under the corresponding assembly action. For cycle time results after compensation, the stable operating reference boundaries are formed using the stable distribution of cycle time results after compensation under the corresponding assembly action. For accompanying changes in the workflow, the stable operating reference boundaries are formed using the stable distributions corresponding to re-inspection changes, waiting changes, and buffer occupancy changes. Preferably, each stable operating reference boundary is formed using the 90th to 95th percentiles of similar stable benchmark samples. When the number of samples is insufficient to form stable percentile boundaries, at least one of the following is used to form stable operating reference boundaries: process cycle time table, quality review procedure, buffer capacity design record, and historical anomaly review samples.

[0041] The degradation overflow identification module generates a compensation proximity result based on the compensation action description quantity. The compensation proximity result is a dimensionless result in the range of 0 to 1. The larger the value, the closer the edge compensation action is to the corresponding calibration compensation upper limit. For compensation occlusion event segments without compensation boundary verification marks, the compensation action description quantity has been occupied and mapped by the compensation occlusion segment construction module according to the calibration compensation upper limit. The degradation overflow identification module uses the compensation action description quantity as the main basis for generating the compensation proximity result. For segments with compensation boundary verification marks, no compensation proximity result is generated due to the lack of an effective calibration compensation upper limit. The execution deviation before compensation, the beat response under the quality release state, and the flow state within the segment boundary are retained for outputting the judgment of occlusion and overflow accompanying changes.

[0042] The degradation spillover identification module forms a compensation dependency result based on a sliding production window. The sliding production window takes the end position of the current compensation occlusion event segment as the window endpoint, and selects consecutive effective assembly action samples forward under the same assembly action. The length of the sliding production window is the dimensionless number of samples, which is jointly determined by the average cycle time of the same stable benchmark samples, the minimum observation time specified in the quality review procedure, and the minimum effective sample size in the process statistics requirements. Preferably, the sliding production window contains 30 to 200 consecutive effective assembly actions; for high-speed stations with a single-piece cycle time of less than 5 seconds, the sliding production window contains 100 to 500 consecutive effective assembly actions; for small batches of workpieces, the sliding production window covers the effective assembly action samples of the three most recent confirmed stable batches. If the historical window samples are insufficient, the observation window is formed by supplementing the samples with stable operation samples confirmed by the process procedure and the review confirmed by the maintenance acceptance record.

[0043] Within the sliding production window, the degradation spillover identification module generates compensation action occurrence results, continuous compensation results, and compensation duration results. The compensation action occurrence results characterize whether the compensation action occurs repeatedly within the window, the continuous compensation results characterize the length of the continuous compensation segment, and the compensation duration results characterize the proportion of the compensation duration to the window time. All of the above results are converted to dimensionless results in the range of 0 to 1, and the larger the value, the higher the degree of compensation dependence. The degradation spillover identification module merges the compensation action occurrence results, continuous compensation results, and compensation duration results into a compensation dependence result according to the weight formed by at least one of the following: historical maintenance review samples, defect attribution records, equipment adjustment records, and process specifications. The weights are non-negative, and the sum of the weights involved in merging within the same sliding production window is 1. The compensation dependence result characterizes the degree of dependence of the corresponding workstation on the cycle response of maintaining the quality release state by the edge compensation action.

[0044] The degradation spillover identification module generates output masking results based on the pre-compensation execution deviation and the post-compensation quality state. The module compares the pre-compensation execution deviation with the corresponding stable operation reference boundary. When the pre-compensation execution deviation exceeds the stable operation reference boundary and the post-compensation quality state meets the release conditions, the corresponding segment is identified as forming an output masking result. The output masking result is a dimensionless result in the range of 0 to 1, used to characterize that the execution deviation has deviated from the stable operation state, but after being absorbed by the edge compensation action, the assembly action still forms the cycle response in the quality release state. If the post-compensation quality state does not meet the release conditions, the segment does not participate in the formation of the output masking result and is retained as a compensation masking failure segment. The compensation masking failure segment is used for maintenance confirmation and recovery observation process.

[0045] The degradation spillover identification module fuses the compensation proximity result, compensation dependency result, and output masking result in the same direction to generate a compensation masking result. Before fusion, the compensation proximity result, compensation dependency result, and output masking result are all dimensionless results in the range of 0 to 1, and the larger the value, the closer it is to the compensation masking state. The fusion weight comes from at least one of the following: historical maintenance review samples, defect attribution records, equipment commissioning records, and process specifications. Preferably, the weight of the output masking result is 0.4 to 0.6, and the remaining weight is distributed between the compensation proximity result and the compensation dependency result, and normalized so that the sum of the three weights is 1. When the number of historical maintenance review samples is insufficient, the weights are determined by the equipment commissioning records and process specifications. This fusion method makes the compensation masking result reflect the state in which the execution deviation has been absorbed by the edge compensation and the quality state after compensation still meets the release conditions, while retaining the constraint effect of the compensation proximity degree and the compensation dependency degree on the gradual degradation trend.

[0046] The conditions for establishing compensation shading are formed by similar stable benchmark samples. Preferably, the 90th to 95th percentiles of the compensation shading results in similar stable benchmark samples are used as the establishment boundary. When the number of similar stable benchmark samples is insufficient, the establishment boundary is jointly determined by the equipment commissioning and acceptance records, stable operation samples confirmed by the process specifications, verification samples confirmed by the quality verification specifications, and verification samples confirmed by the maintenance acceptance records. The degradation spillover identification module determines whether the current compensation shading result meets the compensation shading establishment conditions. If it does, it continues to determine whether the compensation shading result meets the upward trend conditions formed by the comparison of the front and back segments of the sliding production window.

[0047] An upward trend condition is used to determine whether the compensation masking result gradually increases during continuous assembly. Preferably, the upward trend condition is: the median of the compensation masking result in the latter half of the sliding production window is higher than the median of the compensation masking result in the first half, and the difference exceeds the normal fluctuation range of the compensation masking result in the same stable benchmark sample. When the number of samples in the first and second halves is insufficient to support the median comparison, the number of consecutive effective assembly actions determined by the quality review procedure, maintenance review record, or process statistics requirements is used as a supplementary judgment condition. Preferably, the number of consecutive effective assembly actions is 3 to 10. This number is a dimensionless count, and its value is based on the station cycle time, the minimum number of observation actions specified in the quality review procedure, and the number of abnormal confirmation samples in the maintenance review record. By using the upward trend condition, the direct identification of a single compensation fluctuation as a degradation spillover is avoided.

[0048] When the compensated occlusion result meets the conditions for compensation occlusion and satisfies the upward trend condition, the degradation spillover identification module determines the compensated beat offset and accompanying spillover changes. The module compares the compensated beat offset with the corresponding stable operating reference boundary to generate the compensated beat offset. The compensated beat offset is a dimensionless result in the range of 0 to 1; a larger value indicates a higher degree of deviation from the stable operating reference boundary. The degradation spillover identification module also merges the accompanying changes within the segment boundary that deviate from the stable operating reference boundary into spillover accompanying changes. These accompanying changes include... For each of the following changes: re-inspection changes, upstream waiting changes, downstream waiting changes, and buffer usage changes, the degradation spillover identification module first performs a dimensionless conversion according to the corresponding stable operation reference boundary, and then merges them according to the weight determined by at least one of the following: production line cycle balance records, process cycle tables, and historical anomaly review samples, to form the spillover accompanying changes. The weights are non-negative, and the sum of the weights participating in the merging is 1. The spillover accompanying changes are dimensionless results in the range of 0 to 1. The larger the value, the more likely the compensation masking state has been extended to the re-inspection, waiting, or buffer usage level.

[0049] The spillover conditions are formed by at least one of stable production samples, process cycle time tables, quality review procedures, and historical anomaly review samples. Preferably, for the compensated cycle time offset, the spillover conditions use the 90th to 95th percentile of the compensated cycle time result in the same stable benchmark sample as the cycle time offset boundary; for the spillover-accompanying changes, the spillover conditions use the 90th to 95th percentile of the accompanying changes in the flow in the stable production samples as the accompanying change boundary; when the stable production samples are insufficient, the process cycle time table, quality review procedures, buffer capacity design records, and historical anomaly review samples are used to jointly determine the spillover conditions. When at least one of the compensated cycle time offset and the accompanying changes in the spillover meets the spillover conditions, the degradation spillover identification module generates a degradation spillover segment. This judgment method ensures that the degradation spillover segment has both compensation occlusion results and spillover state support, avoiding the generation of degradation spillover segments due to a single compensation action or a single cycle time fluctuation.

[0050] The generated degraded overflow fragments include the corresponding compensated occlusion event fragments, the compensated occlusion result, the compensated beat offset, the overflow-related changes, the rising state of the compensated occlusion result, the edge state version, and the fragment time boundary. The degraded overflow fragments are used as the starting point for subsequent propagation by the propagation attribution module to compensated constraint attribution of the propagation relationship between the degraded overflow fragments and the affected fragments. This output enables subsequent propagation analysis to start from the fragments that have formed compensated occlusion and have an overflow state, rather than starting directly from ordinary beat fluctuations or ordinary buffer occupancy changes.

[0051] In this embodiment, the propagation attribution module receives the degradation spillover fragments generated by the degradation spillover identification module, uses the degradation spillover fragments as the propagation starting point, determines the affected fragments according to the validity of the production line flow sequence and the reliability of the cloud edge status, and performs compensation constraint attribution on the propagation relationship between the degradation spillover fragments and the affected fragments to form a compensation constraint beat propagation relationship diagram and generate degradation anomaly source tags. This step is used to distinguish the beat propagation caused by degradation spillover from non-degradation anomalies caused by workpiece model switching, incoming material fluctuations, downstream congestion, buffer occupation, or data upload latency.

[0052] The propagation attribution module establishes the production line flow path based on the production line process route, station connection relationship, buffer area connection relationship, and workpiece flow record. The production line flow path serves as the constraint input for compensation constraint attribution, representing the directional relationship of a workpiece entering the current station from the upstream station and continuing to flow to the downstream station. For connection relationships with buffer areas, transfer mechanisms, or AGVs, the production line flow path also includes buffer entry time, buffer release time, transfer arrival time, and workpiece arrival time. The production line flow path is used to determine the propagation observation range, perform propagation direction verification, and filter affected segments.

[0053] The propagation attribution module uses the degraded spillover fragment as the propagation starting point and determines the propagation observation range based on the production line flow path. The propagation observation range is formed by the standard flow time, buffer release time, and maximum allowable waiting time of the production line. The standard flow time, buffer release time, and maximum allowable waiting time of the production line all have time dimensions, and the units are consistent with the time units after time base correction in the compensation masking fragment construction module. The standard flow time is derived from at least one of the process cycle table, transfer mechanism records, and historical stable production samples. Preferably, the 5th to 95th percentiles of the flow time in historical stable production samples under the same process route, the same workpiece model, and the same transfer method are used to form the lower and upper bounds of the flow time. The buffer release time is derived from the buffer design records and historical data. At least one of the stable production samples is used, preferably the buffer release time is the 90th to 95th percentile of the buffer release time in the historical stable production samples, to form the buffer release boundary; the maximum allowable waiting time of the production line is derived from at least one of the process cycle balance record, buffer capacity design record, and production scheduling procedure. The propagation attribution module takes the fragment time boundary of the degraded spillover fragment as the starting position, combines the lower bound of the flow time to determine the earliest occurrence boundary of the affected fragment, and combines the upper bound of the flow time, the buffer release boundary, and the maximum allowable waiting time of the production line to determine the latest occurrence boundary of the affected fragment, forming the propagation observation range; if the historical stable production samples are insufficient, the process cycle table, the transfer mechanism acceptance record, and the buffer capacity design record are used together to determine the propagation observation range.

[0054] Within the propagation observation range, the propagation attribution module identifies affected segments. Affected segments refer to segments located in the adjacent upstream stations, adjacent downstream stations, extended stations connected via buffer areas, or extended stations connected via transfer mechanisms of the station corresponding to the degraded overflow segment, where waiting changes, arrival delays, cycle time offsets, buffer occupancy changes, or re-inspection changes occur. Affected segments are matched based on segment time boundaries, assembly action attribution, and edge state version. When a candidate affected segment does not belong to the same workpiece flow link as the degraded overflow segment, or its time change is not within the propagation observation range, it is not identified as an affected segment. Affected segments serve as inputs for flow timing verification, propagation direction verification, compensation overflow association verification, and cloud-edge timing reliability verification.

[0055] The propagation attribution module performs flow timing verification, propagation direction verification, compensation overflow association verification, and cloud-edge timing reliability verification on the propagation relationship between degraded overflow fragments and affected fragments. Flow timing verification is used to confirm that the affected fragment is within the propagation observation range. When the waiting change, arrival delay, buffer occupancy change, or cycle offset of the affected fragment occurs earlier than the fragment time boundary of the degraded overflow fragment, or exceeds the maximum allowable waiting time of the production line, the corresponding propagation relationship will not pass the flow timing verification. For scenarios with buffer areas or transfer equipment, the buffer release time, transfer arrival time, and workpiece arrival time are included in the flow timing verification. The flow timing verification result is a dimensionless result in the range of 0 to 1. The larger the value, the more consistent the change time of the affected fragment is with the propagation observation range.

[0056] The propagation direction verification is used to confirm that the direction of state change matches the production line flow path. When a degenerate overflow fragment is located at the current station and the downstream station experiences arrival delay, waiting increase, buffer release delay, or cycle offset, the propagation direction verification determines whether the change is consistent with the workpiece flow direction. When the upstream station experiences waiting accumulation, the propagation direction verification combines the buffer occupancy status and the downstream blockage release time to determine whether the change is a reverse propagation caused by the downstream blockage. If the change direction of the candidate affected fragment is inconsistent with the production line flow path and cannot be explained by buffer occupancy, transfer delay, or workpiece stagnation position, no compensation constraint propagation edge is established. The propagation direction verification result is a dimensionless result in the range of 0 to 1. The larger the value, the more consistent the direction of state change is with the production line flow path.

[0057] The compensation overflow correlation check is used to confirm that the changes in the affected segment are supported by the compensation masking result, the compensated beat offset, and the accompanying overflow changes in the degraded overflow segment. The degraded overflow segment should have a compensation masking result and at least one of the compensated beat offset and the accompanying overflow changes. The waiting changes, beat offsets, buffer occupancy changes, or re-inspection changes in the affected segment should be able to form a corresponding relationship with the compensation masking result, the compensated beat offset, and the accompanying overflow changes in the degraded overflow segment within the propagation observation range. If the degraded overflow segment only has a single compensation fluctuation, does not form a compensation masking result, and does not trigger the compensated beat offset or the accompanying overflow changes, then it will not pass the compensation overflow correlation check. The compensation overflow correlation check result is a dimensionless result in the range of 0 to 1. The larger the value, the more the changes in the affected segment are supported by the degraded overflow segment.

[0058] The cloud-edge time-series reliability verification is used to exclude propagation relationships formed solely by cloud-received time sequences. The propagation attribution module uses the original generation time at the edge as the first time sequence criterion and the cloud-received time sequence as the second time sequence criterion, comparing the chronological order of degraded spillover segments and affected segments. Only when the propagation chronological order is valid in the original generation time sequence at the edge and the cloud-received time sequence does not change the propagation direction is the propagation relationship allowed to continue participating in the establishment of compensation constraint propagation edges. If the propagation chronological order is valid only in the cloud-received time sequence but not in the original generation time sequence at the edge, the relationship is attributed to the data. The non-degenerate anomaly source label corresponding to the transmission delay does not establish a compensation constraint propagation edge. The stable boundary of the data upload delay comes from at least one of the cloud-edge communication logs, edge gateway clock synchronization records, and historical stable upload records. Preferably, the 90th to 95th percentile of the historical stable upload delay is used as the delay interpretation boundary. When there are insufficient historical upload records, the edge gateway acceptance records and cloud-edge communication procedures are used to determine the delay interpretation boundary. The cloud-edge timing reliability verification result is a dimensionless result in the range of 0 to 1. The larger the value, the less the propagation order is affected by the data upload delay.

[0059] When the propagation relationship between the degraded spillover fragment and the affected fragment passes the flow timing verification, propagation direction verification, compensation spillover association verification, and cloud-edge timing reliability verification, and is not fully interpreted by the non-degraded anomaly source label, the propagation attribution module establishes a compensation constraint propagation edge. The compensation constraint propagation edge is used to connect the degraded spillover fragment and the affected fragment, and records the verification type, propagation observation range, edge state version, and fragment time boundary. The compensation constraint propagation edge is a constrained propagation relationship that simultaneously satisfies the production line flow timing, propagation direction, compensation spillover association, and cloud-edge timing reliability. Simple workstation adjacency relationship is not used as the basis for establishing the compensation constraint propagation edge.

[0060] The propagation attribution module verifies the basis for the formation of degenerate anomaly source tags based on non-degenerate anomaly source tags. Non-degenerate anomaly source tags are used to characterize the explanatory relationship of non-degenerate causes to cycle time changes. Non-degenerate causes include at least one of workpiece model switching, incoming material fluctuations, downstream congestion, buffer occupancy, and data upload latency. The propagation attribution module performs a coverage-based interpretation verification on non-degenerate anomaly source tags: when a non-degenerate anomaly source tag can cover all anomaly changes of the same event anchor point, the same state segment, and the same propagation edge, a degenerate anomaly source tag is not generated; when a non-degenerate anomaly source tag only explains part of the propagation edge, only the explained propagation edge is excluded, and the unexplained compensation constraint propagation edge is retained. For the same candidate propagation relationship, only one exclusion rule is used as the main exclusion basis, and the other exclusion rules are used as auxiliary explanations, without repeatedly weakening the basis for the formation of degenerate anomaly source tags. This processing is used to avoid repeatedly excluding the same degenerate propagation relationship when multiple non-degenerate causes occur at the same time.

[0061] When both non-degenerate anomaly source tags and degenerate anomaly source tags exist for the same affected segment, the propagation attribution module first determines whether the non-degenerate anomaly source tag covers all anomaly changes of the same event anchor point, the same state segment, and the same compensation constraint propagation edge. Only when all anomaly changes are covered is the basis for the formation of the degenerate anomaly source tag corresponding to the propagation edge excluded. If the non-degenerate anomaly source tag only covers some changed objects, some time intervals, or some propagation edges, then only the covered part is excluded. The compensation constraint propagation edge that is not covered continues to participate in the propagation intensity result and the formation of the degenerate anomaly source tag. This process is used to avoid mixing workpiece model switching, incoming material fluctuations, downstream congestion, cache occupation, or data upload latency with the actual degenerate spillover propagation.

[0062] The propagation attribution module constructs a compensation constraint beat propagation relationship graph based on degenerate spillover segments, affected segments, and compensation constraint propagation edges. Nodes in the graph include degenerate spillover segments and affected segments connected by compensation constraint propagation edges. Edges include propagation relationships that satisfy the conditions for establishing compensation constraint propagation edges. The propagation attribution module aligns the flow timing verification results, propagation direction verification results, compensation spillover association verification results, and cloud-edge timing reliability verification results, then performs a weighted merging to form a propagation strength result. The propagation strength result is a dimensionless result in the range of 0 to 1. A larger value indicates that the compensation constraint propagation edge can better support the propagation interpretation of affected segments by degenerate spillover segments. The weights of the propagation strength result are derived from at least one of the following: production line beat balance records, historical anomaly review samples, and key workstation calibration records. The weights are non-negative and the sum of the weights is 1. If historical anomaly review samples are insufficient, production line beat balance records and key workstation calibration records are used to determine the weights.

[0063] The conditions for establishing propagation intensity are formed by at least one of the following: ordinary flow fluctuation samples, confirmed degradation propagation samples, production line cycle balance records, and key workstation calibration records. Ordinary flow fluctuation samples are derived from historical stable production samples where no degradation spillover has occurred and the quality release status is stable. Confirmed degradation propagation samples are derived from historical anomaly review samples and maintenance confirmation records. Both ordinary flow fluctuation samples and confirmed degradation propagation samples are sets of propagation edge samples, and their sample count is dimensionless. Preferably, each sample set contains no less than 30 propagation edge samples. For high-speed workstations or strongly coupled buffer production lines with a cycle time of less than 5 seconds, each sample set preferably contains no less than 100 propagation edge samples. If the sample count is insufficient, production line cycle balance records are used in conjunction with the samples. The criteria for determining the establishment boundary are as follows: 1. Key workstation calibration records and maintenance confirmation records. Preferably, the criteria for establishing the propagation intensity use the 90th to 95th percentiles of the propagation intensity of ordinary flow fluctuation samples as the ordinary fluctuation exclusion boundary, and the 10th to 25th percentiles of the propagation intensity of confirmed degradation propagation samples as the degradation propagation retention boundary. When the ordinary fluctuation exclusion boundary is lower than the degradation propagation retention boundary, an establishment boundary is determined between the ordinary fluctuation exclusion boundary and the degradation propagation retention boundary. When the two overlap or intersect, the establishment boundary is determined by combining the production line cycle balance records, key workstation calibration records, and maintenance confirmation records. This condition is used to exclude ordinary flow fluctuations while avoiding filtering out moderate-intensity but real degradation propagation.

[0064] When a degenerate spillover fragment has a compensation occlusion result, at least one compensation constraint propagation edge, and the propagation intensity result meets the propagation intensity condition, the propagation attribution module generates a degenerate anomaly source label. The degenerate anomaly source label includes the degenerate spillover fragment, the set of affected fragments, the set of compensation constraint propagation edges, the compensation occlusion result, the beat offset after compensation, the accompanying changes of spillover, the propagation intensity result, and the edge state version. This label is used by the subsequent cloud-edge control verification module to generate edge anomaly state records, determine the state influence range, and perform state valid interval verification.

[0065] For abnormal segments that do not meet the conditions for forming a degraded anomaly source label, the propagation attribution module generates corresponding non-degraded anomaly source labels according to the non-degraded causes that are fully explained. These include model switching type cycle time anomaly labels, incoming material fluctuation type cycle time anomaly labels, downstream blockage type cycle time anomaly labels, buffer occupancy type cycle time anomaly labels, and data upload delay type cycle time anomaly labels. The above non-degraded anomaly source labels are used to explain the non-degraded causes of the corresponding cycle time changes and are not output as degraded anomaly source labels.

[0066] In this embodiment, the cloud-edge control verification module receives the degradation anomaly source label and compensation constraint beat propagation relationship diagram generated by the propagation attribution module, and verifies the effective range of the execution state of the cloud scheduling instruction in the pre-execution state in combination with the edge state version, generating a scheduling verification result. The scheduling verification result includes at least one of the workstation scheduling restriction result and the scheduling instruction to be reviewed. When the recovery closure result formed by the recovery observation is valid, a scheduling cancellation result and anomaly rollback benefit result are generated. This step makes cloud scheduling, maintenance confirmation and edge state update jointly constrained by the degradation anomaly source label, state influence range, edge state version and edge state effective range.

[0067] The cloud-edge control verification module generates edge anomaly state records based on the degenerate anomaly source label and the compensation constraint beat propagation relationship graph. The edge anomaly state records include the degenerate anomaly source label, compensation occlusion result, degenerate spillover fragment, compensation constraint propagation edge set, propagation intensity result, edge state version, and state generation time. Among them, the degenerate anomaly source label serves as the basis for judging the anomaly source, the compensation constraint propagation edge set serves as the basis for determining the state's influence range, and the edge state version serves as the basis for the version of the anomaly state referenced by the cloud-based scheduling instructions to be executed.

[0068] The cloud-edge control verification module determines the state influence range based on the compensation constraint propagation edge. The state influence range includes the workstation corresponding to the degenerate overflow segment, as well as the workstation corresponding to the affected segment connected through the compensation constraint propagation edge. When the compensation constraint propagation edge is connected to the buffer area, transfer mechanism, or AGV connected workstation, the state influence range also includes the workstation range associated with the corresponding buffer area, transfer mechanism, or AGV. The state influence range is bounded by the node actually connected by the compensation constraint propagation edge and does not extend to workstations not connected by the compensation constraint propagation edge.

[0069] The cloud-edge control verification module forms an effective range of edge states based on the generation time of the degraded spillover fragment, the generation time of the degraded anomaly source label, and the natural expiration time of the state. After the recovery closure result is generated, the end boundary of the effective range of edge states is updated with the generation time of the recovery closure result. For the trigger verification of the scheduling instructions to be executed in the cloud, the generation time of the degraded anomaly source label is used as the starting point for the state reference. For the attribution of the anomaly rollback benefit result, the generation time of the degraded spillover fragment is used as the starting point for the state influence, and the natural expiration time of the state is the time quantity, which is formed by the propagation observation range, the maximum allowable waiting time of the production line, and the cycle response boundary in the stable operation reference boundary. Preferably, the later of the following times is used: the end time of the propagation observation range, the expiration time of the maximum allowable waiting time of the production line, and the time required to complete the recovery observation according to the stable operation reference boundary. When historical samples are insufficient, the natural expiration time of the state is determined by the process cycle table, quality review procedure, buffer capacity design record, and maintenance acceptance record. When the same degraded anomaly source label is generated again before the natural expiration time of the state, the generation time of the latest degraded anomaly source label is used to update the effective range of edge states.

[0070] The cloud-edge control verification module reads the edge state version, planned execution time, and workstation range referenced by the scheduling instruction to be executed in the cloud, and performs a valid state interval verification. When the edge state version referenced by the scheduling instruction to be executed in the cloud is consistent with the edge state version in the edge abnormal state record, the planned execution time of the instruction falls within the valid edge state interval, and there is an intersection between the workstation range of the instruction and the state influence range, a workstation scheduling restriction result is generated. The workstation scheduling restriction result includes at least one of the following: speed limit, postponement of order insertion, restriction of parallel dispatch, maintenance confirmation prompt, and quality review prompt for workstations within the state influence range, and only applies to workstations within the state influence range.

[0071] When the edge state version referenced by the cloud-based scheduling instruction to be executed is inconsistent with the edge state version in the edge abnormal state record, or when the scheduled execution time of the instruction is not within the effective range of the edge state, or when there is no intersection between the workstation range affected by the instruction and the range of state influence, the cloud-edge control verification module will not generate workstation scheduling restriction results and will mark the cloud-based scheduling instruction to be executed as a scheduling instruction to be reviewed. The scheduling instruction to be reviewed is used to enable the cloud to re-acquire the edge state version, the range of state influence, and the effective range of the edge state, so as to avoid continuing to execute rate limiting, rescheduling, line stop, or maintenance triggers based on the old version state.

[0072] The cloud-edge control verification module configures processing action tags for compensation occlusion event segments in the recovery observation. These tags include action type, action execution time, affected workstation range, corresponding edge state version, the degenerate anomaly source tag that triggered the action, and the corresponding compensation constraint propagation edge. Action types include at least one of workstation scheduling restrictions, maintenance confirmation, and pending review processing. The cloud-edge control verification module only includes a state change as an endogenous state change in the anomaly rollback benefit result if the state change in the recovery observation segment occurs after the processing action execution time and is within the affected workstation range or the state's influence range. If the state change occurs before the processing action, or... If the change occurs outside the scope of the state's influence, it will be recorded as an external disturbance or a non-target recovery change and will not be included in the attribution of abnormal rollback benefits. The processing action flag is used to distinguish between endogenous state changes and external disturbances caused by at least one of workstation scheduling restrictions, maintenance confirmations, and pending review processes. During the recovery observation period, if there is a drop in compensation occlusion results, a drop in post-compensation cycle offset, or the removal of overflow-related changes, the cloud-edge control verification module will determine whether the change occurred after at least one of workstation scheduling restrictions, maintenance confirmations, and pending review processes based on the processing action flag. If the state change occurs before the processing action, or if the corresponding workstation is not within the scope of the state's influence, it will not be regarded as a recovery change caused by the processing action of this system.

[0073] During the recovery observation process, the cloud-edge control verification module takes the state influence range as the observation object and obtains recovery observation segments. The recovery observation segments include compensation occlusion event segments, compensation occlusion results, post-compensation cycle time offset, overflow accompanying changes, quality release status, and edge status versions within the state influence range. The recovery observation window is jointly determined by the quality review procedure, maintenance acceptance records, and the average cycle time of similar stable benchmark samples. The number of samples in the recovery observation window is the number of effective assembly actions, which is a dimensionless count. Preferably, the recovery observation window includes no less than 10 effective assembly actions. For high-speed stations with a single-piece cycle time of less than 5 seconds, the recovery observation window includes no less than 50 effective assembly actions. For small-batch or multi-model mixed-line scenarios, the recovery observation window covers at least one stable batch that has completed quality review. When historical stable samples are insufficient, the recovery observation window is determined using the quality review procedure, maintenance acceptance records, and process cycle time table.

[0074] The cloud-edge control verification module forms a recovery benchmark based on at least one of the following: similar stable benchmark samples, maintenance acceptance records, and quality review procedures. The recovery benchmarks correspond to the compensation occlusion results, the compensation cycle offset, and the accompanying overflow changes, respectively. Different recovery benchmarks are formed separately without direct merging across dimensions. For the compensation occlusion results, the recovery benchmark is formed based on the stable distribution of compensation occlusion results in similar stable benchmark samples. For the compensation cycle offset, the recovery benchmark is formed based on the cycle response boundary in the stable operation reference boundary. For the accompanying overflow changes, the recovery benchmark is formed based on the stable distribution corresponding to the re-inspection changes, waiting changes, and buffer occupancy changes, respectively. Preferably, the recovery benchmark adopts a stable boundary no higher than the 90th percentile of similar stable benchmark samples, or adopts the acceptance boundary confirmed by the maintenance acceptance records and quality review procedures. When historical samples are insufficient, the recovery benchmark is jointly determined by the process procedures, maintenance acceptance records, and quality review procedures.

[0075] When the compensation occlusion result, the compensated cycle offset, and the accompanying changes in overflow all return to the recovery baseline and meet the continuous quality release conditions, and the recovery status version synchronization is completed, the cloud-edge control verification module determines that the recovery closure verification has passed and generates the recovery closure result. The continuous quality release conditions are formed by at least one of the quality review procedures and maintenance acceptance records, and are used to confirm that the assembly actions within the state influence range maintain the quality release status within the recovery observation window. The sample number in the continuous quality release conditions is the number of valid assembly actions, which is a dimensionless count; preferably, the continuous quality release conditions include 10 to 5 consecutive Zero valid assembly actions meet the quality release requirements; for high-speed workstations with a single-piece cycle time of less than 5 seconds, the continuous quality release condition includes 50 to 200 consecutive valid assembly actions meeting the quality release requirements; for small-batch or multi-model mixed-line scenarios, the continuous quality release condition includes at least one stable batch passing the quality review. The above values ​​are determined based on the quality review procedure, maintenance and acceptance records, the average cycle time of similar stable benchmark samples, and process acceptance requirements. Recovery status version synchronization refers to the edge side updating the edge status version within the scope of status influence to the recovery status version and synchronizing the recovery status version to the cloud.

[0076] When the recovery closure verification fails, the cloud-edge control verification module does not generate a scheduling cancellation result; if the compensation occlusion result is still higher than the recovery benchmark, the compensation cycle offset is still higher than the recovery benchmark, or the overflow-accompanied change has not been resolved, the edge abnormal state record and the corresponding workstation scheduling restriction result are maintained; if a new degenerate anomaly source label appears in the recovery observation segment, the effective range of the edge state and the state influence range are updated based on the new degenerate anomaly source label.

[0077] When the recovery closure verification passes, the cloud-edge control verification module generates a scheduling cancellation result. The scheduling cancellation result includes the cancelled workstation scheduling restriction result, the corresponding degradation anomaly source label, the recovery status version, the time when the recovery closure result is generated, and the scope of cancellation. Based on the scheduling cancellation result, the cloud-edge control verification module cancels the speed limit, delayed order insertion, maintenance confirmation prompt, or quality review prompt for workstations within the status influence range, and transfers unexecuted scheduling instructions referencing the old edge status version to the pending review scheduling instructions. For workstations outside the status influence range, no scheduling cancellation result is generated.

[0078] After generating the scheduling cancellation result, the cloud-edge control verification module generates the abnormal rollback benefit result. The abnormal rollback benefit result is bound to the degradation anomaly source label, the state impact range, the compensation constraint propagation edge, and the scheduling cancellation result. It is used to characterize the effect of at least one of the scheduling restrictions, maintenance confirmation, and pending review processing on the removal of the degradation spillover state. The abnormal rollback benefit result includes the fallback of the compensation occlusion result, the fallback of the clock offset after compensation, the removal of the spillover-accompanying change, the continuous quality release, the recovery state version synchronization, and the scope of the scheduling cancellation. The abnormal rollback benefit result is not allocated according to a fixed priority, but is bound to the degradation anomaly source label, the state impact range, and the compensation constraint propagation edge, so that the benefit attribution corresponds to the real change object and propagation relationship.

[0079] Finally, the cloud-edge control verification module outputs the production line automation management results, which include edge abnormal status records, edge status effective range, status impact range, workstation scheduling restriction results, scheduling instructions to be reviewed, processing action markers, recovery closure results, scheduling cancellation results, and abnormal rollback benefit results. This output is used to ensure that cloud scheduling, maintenance confirmation, and edge status updates are executed in a closed loop within the same edge status version and status effective range.

[0080] Example 2: Please see Figure 2 Based on Example 1, this embodiment also provides an automated management method for an assembly workpiece production line, including the following steps: By accessing assembly operation observation data and cloud-edge scheduling status data, and taking the edge compensation trigger point that completes time base correction as the event anchor point, the execution status before and after compensation is locked according to the assembly action affiliation and edge status version. A compensation occlusion closure relationship is constructed to represent the execution deviation being absorbed by edge compensation and forming the beat response in the quality release state, and compensation occlusion event fragments are generated. Perform deviation in the same direction conversion, compensation occupancy characterization, and quality release comparison with stable beat on the compensation occupancy event segment to generate compensation occupancy results. When the compensation occupancy result is in an upward state within the sliding production window and triggers beat overflow state, generate degenerate overflow segment. Starting from the degraded spillover fragments, the affected fragments are determined according to the validity of the production line flow sequence and the credibility of the cloud edge status. The propagation relationship between the degraded spillover fragments and the affected fragments is attributed with compensation constraints, forming a compensation constraint beat propagation relationship diagram, and generating a degraded anomaly source label. Based on the degenerate anomaly source label, the compensation constraint beat propagation relationship diagram, and the edge state version, the effective interval of the execution state of the cloud scheduling instruction in the pre-execution state is verified, and a scheduling verification result is generated. The scheduling verification result includes at least one of the workstation scheduling restriction result and the scheduling instruction to be reviewed. When the recovery closure result formed by the recovery observation is valid, a scheduling cancellation result and anomaly rollback benefit result are generated.

[0081] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0082] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0083] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. An automated management system for an assembly line, characterized in that, include: The compensation occlusion segment construction module connects assembly operation observation data and cloud-edge scheduling status data. It uses the edge compensation trigger point that completes time base correction as the event anchor point, locks the execution status before and after compensation according to the assembly action affiliation and edge status version, constructs a compensation occlusion closure relationship that represents the absorption of execution deviation by edge compensation, and forms the beat response under the quality release state, generating compensation occlusion event segments. The degradation overflow identification module performs deviation in the same direction conversion, compensation occupancy characterization, and quality release comparison with stable beat for the compensation occupancy event segment, generates compensation occupancy results, and generates degradation overflow segments when the compensation occupancy results are in an upward state within the sliding production window and trigger beat overflow state. The propagation attribution module takes the degraded spillover fragment as the propagation starting point, determines the affected fragment according to the validity of the production line flow sequence and the credibility of the cloud edge state, and performs compensation constraint attribution on the propagation relationship between the degraded spillover fragment and the affected fragment, forming a compensation constraint beat propagation relationship diagram and generating degraded anomaly source labels. The cloud-edge control verification module verifies the effective range of the execution state of cloud scheduling instructions in the pre-execution state based on the degradation anomaly source label, the compensation constraint beat propagation relationship diagram, and the edge state version. It generates scheduling verification results, which include at least one of the workstation scheduling restriction results and scheduling instructions to be reviewed. When the recovery closure result formed by the recovery observation is valid, it generates scheduling cancellation results and anomaly rollback benefit results.

2. The automated management system for an assembly line of workpieces according to claim 1, characterized in that, Using the edge compensation trigger point that has completed time base correction as the event anchor point, the process includes: performing time base correction on assembly operation observation data and cloud edge scheduling status data based on at least one of edge node clock synchronization records, cloud-edge communication logs, and a unified clock source for the production line; using the edge compensation trigger point after time base correction as the segment start boundary, the execution closing time of the same assembly action as the segment first closing boundary, and the quality inspection result return time as the segment second closing boundary; performing consistency matching on the segment start boundary, segment first closing boundary, and segment second closing boundary according to the assembly action attribution and edge state version to form segment boundaries; when the quality inspection result has not returned, the corresponding segment is marked as an unclosed segment, and delayed closure is performed according to the assembly action attribution and edge state version after the quality inspection result returns.

3. The automated management system for an assembly line of workpieces according to claim 2, characterized in that, The process of forming the compensation masking closure relationship of the cycle response under the quality release state includes: within the segment boundary, performing a dimensionless conversion based on the corresponding execution reference and deviation boundary on the execution feedback quantity before compensation triggering to obtain the execution deviation before compensation; performing an occupancy mapping based on the corresponding calibration compensation upper limit on the edge control compensation record to obtain the compensation action description quantity; when the quality state after compensation meets the release conditions, mapping the cycle result after compensation to the cycle response under the quality release state, and closing the execution deviation before compensation, the compensation action description quantity, and the cycle response under the quality release state according to the segment boundary to generate a compensation masking event segment used to characterize the execution deviation being absorbed by edge compensation; wherein, the execution reference, deviation boundary, and calibration compensation upper limit are derived from equipment calibration records, process specifications, commissioning and acceptance records, or historical stable production samples.

4. The automated management system for an assembly line of workpieces according to claim 3, characterized in that, The compensation occlusion segment construction module is also used to configure segment status markers for compensation occlusion event segments. Segment status markers include closed segment markers, unclosed segment markers, version to be verified markers, and compensation boundary to be verified markers. When the edge state version referenced by the cloud scheduling instruction is inconsistent with the edge state version when the assembly action is actually executed, a version to be verified marker is configured. The version to be verified marker is used to restrict the direct execution of the cloud scheduling instruction. When the compensation parameter does not have a calibrated compensation upper limit or the calibrated compensation upper limit is zero, a compensation boundary to be verified marker is configured. Compensation occlusion event segments with compensation boundary to be verified markers do not participate in subsequent degradation overflow identification. Based on the calibrated compensation upper limit, the processing of compensation approximation results is formed. The execution deviation before compensation, the beat response under the quality release state, and the flow state within the segment boundary are retained as the basis for judging output occlusion, overflow accompanying changes, and boundary verification.

5. The automated management system for an assembly line of workpieces according to claim 4, characterized in that, The operation for generating compensation masking results includes: establishing a set of similar stable benchmark samples according to the assembly action affiliation; forming a stable operation reference boundary from the set of similar stable benchmark samples for compensation masking identification and degradation spillover identification; forming compensation proximity results based on the compensation action description quantity; forming compensation dependency results based on whether the compensation action occurs within the sliding production window, the continuous compensation state, and the compensation duration state; forming output masking results based on segments where the execution deviation before compensation exceeds the stable operation reference boundary and the quality state after compensation meets the release conditions; and merging the compensation proximity results, compensation dependency results, and output masking results in the same direction to generate the compensation masking result. The set of similar stable benchmark samples is formed from sample records confirmed by production stability. These sample records include equipment commissioning and acceptance records, continuous stable production cycle records, stable operation samples confirmed by process procedures, verification-passed samples confirmed by quality verification procedures, and verification-passed samples confirmed by maintenance acceptance records.

6. The automated management system for an assembly line according to claim 5, characterized in that, The operation of generating degraded spillover segments includes: when the compensation masking result meets the compensation masking establishment condition formed by similar stable benchmark samples, and the compensation masking result meets the upward trend condition formed by the comparison of the front and rear segments of the sliding production window, the compensated cycle time result is compared with the stable operation reference boundary, and the flow-related changes that deviate from the stable operation reference boundary within the segment boundary are merged into spillover-related changes; a cycle time spillover state is formed based on the compensated cycle time offset and spillover-related changes; when the cycle time spillover state meets the spillover establishment condition formed by at least one of the stable production samples, process cycle time table, quality review procedure, and historical anomaly review samples, a degraded spillover segment is generated; segments whose quality status after compensation does not meet the release conditions do not participate in the formation of the output masking result, but are retained as compensation masking failure segments, which are used for maintenance confirmation and recovery observation processes.

7. The automated management system for an assembly line of workpieces according to claim 6, characterized in that, The operation of compensating for the attribution of the propagation relationship between the degraded spillover fragment and the affected fragment includes: taking the degraded spillover fragment as the propagation starting point, determining the propagation observation range based on the production line flow path, and identifying the affected fragment within the propagation observation range; performing flow timing verification, propagation direction verification, compensation spillover correlation verification, and cloud-edge timing reliability verification on the propagation relationship between the degraded spillover fragment and the affected fragment; flow timing verification is used to confirm that the affected fragment is within the propagation observation range, propagation direction verification is used to confirm that the direction of state change matches the production line flow path, and compensation spillover correlation verification is used to confirm that the change of the affected fragment is supported by the compensation occlusion result and the cycle spillover state in the degraded spillover fragment. Cloud-edge time sequence reliability verification is used to exclude propagation relationships formed solely by cloud-received time sequences. When a propagation relationship passes verification and is not fully explained by non-degenerate anomaly source labels, a compensating constraint propagation edge is established. Based on the compensating constraint propagation edge, a compensating constraint beat propagation relationship graph and degenerate anomaly source labels are formed. Non-degenerate anomaly source labels are used to characterize the explanatory relationship of non-degenerate causes on beat changes. The propagation observation range is formed by standard flow time, buffer release time, and maximum allowable waiting time of the production line. The standard flow time, buffer release time, and maximum allowable waiting time of the production line are derived from at least one of the following: process beat table, buffer design record, transfer mechanism record, and historical stable production sample.

8. The automated management system for an assembly line according to claim 7, characterized in that, The operation of verifying the basis for forming degenerate anomaly source tags based on non-degenerate anomaly source tags includes: performing a comprehensive interpretation verification on non-degenerate anomaly source tags; when a non-degenerate anomaly source tag can cover all anomaly changes of the same event anchor point, the same state segment, and the same propagation edge, no degenerate anomaly source tag is generated; when a non-degenerate anomaly source tag only interprets part of the propagation edge, only the interpreted propagation edge is excluded, and the uninterpreted compensation constraint propagation edge is retained; for the same candidate propagation relationship, only one exclusion rule is used as the main exclusion basis, and the other exclusion rules are used as auxiliary explanations, without repeatedly weakening the basis for forming degenerate anomaly source tags; wherein, non-degenerate anomaly source tags include tags used to characterize at least one of the following interpretation relationships of cycle time changes: workpiece model switching, incoming material fluctuation, downstream congestion, buffer occupation, and data upload latency.

9. The automated management system for an assembly line of workpieces according to claim 8, characterized in that, The operations for generating recovery closure results, scheduling cancellation results, and anomaly rollback benefit results include: generating edge anomaly state records based on the degenerate anomaly source label and the compensation constraint beat propagation relationship diagram; determining the state influence range based on the compensation constraint propagation edge; forming the edge state effective interval based on the degenerate spillover fragment generation time, the degenerate anomaly source label generation time, and the state natural failure time; updating the end boundary of the edge state effective interval with the recovery closure result generation time after the recovery closure result is generated; wherein, the state natural failure time is formed by the propagation observation range, the maximum allowable waiting time of the production line, and the beat response boundary in the stable operation reference boundary; verifying the edge state version, planned execution time, and effective workstation range referenced by the cloud scheduling instruction in the pre-execution state, generating scheduling verification results, which include workstation scheduling restriction results and the schedule to be reviewed. At least one of the degree instructions is used to configure processing action markers for the compensation occlusion event segments in subsequent recovery observations. The processing action markers are used to distinguish between endogenous state changes and external disturbances caused by at least one of workstation scheduling restrictions, maintenance confirmations, and pending review processing. During the recovery observation process, the compensation occlusion state, cycle overflow state, and flow-accompanying state all revert to the recovery benchmark formed by similar stable benchmark samples, maintenance acceptance records, or quality review procedures, satisfying the continuous quality release conditions formed by at least one of the quality review procedures and maintenance acceptance records. At the same time, the recovery state version synchronization is completed, serving as the recovery closure verification condition. When the recovery closure verification passes, a recovery closure result is generated. Based on the recovery closure result, a scheduling cancellation result and anomaly rollback benefit result are generated. The anomaly rollback benefit result is bound to the degradation anomaly source label, state influence range, compensation constraint propagation edge, and scheduling cancellation result.

10. An automated management method for an assembly line, employing an automated management system for an assembly line as described in any one of claims 1-9, characterized in that, Includes the following steps: By accessing assembly operation observation data and cloud-edge scheduling status data, and taking the edge compensation trigger point that completes time base correction as the event anchor point, the execution status before and after compensation is locked according to the assembly action affiliation and edge status version. A compensation occlusion closure relationship is constructed to represent the execution deviation being absorbed by edge compensation and forming the beat response in the quality release state, and compensation occlusion event fragments are generated. Perform deviation in the same direction conversion, compensation occupancy characterization, and quality release comparison with stable beat on the compensation occupancy event segment to generate compensation occupancy results. When the compensation occupancy result is in an upward state within the sliding production window and triggers beat overflow state, generate degenerate overflow segment. Starting from the degraded spillover fragments, the affected fragments are determined according to the validity of the production line flow sequence and the credibility of the cloud edge status. The propagation relationship between the degraded spillover fragments and the affected fragments is attributed with compensation constraints, forming a compensation constraint beat propagation relationship diagram, and generating a degraded anomaly source label. Based on the degenerate anomaly source label, the compensation constraint beat propagation relationship diagram, and the edge state version, the effective interval of the execution state of the cloud scheduling instruction in the pre-execution state is verified, and a scheduling verification result is generated. The scheduling verification result includes at least one of the workstation scheduling restriction result and the scheduling instruction to be reviewed. When the recovery closure result formed by the recovery observation is valid, a scheduling cancellation result and anomaly rollback benefit result are generated.

Images