A device manufacturing state tracing method and system based on internet of things collection

Abstract

The present application belongs to the technical field of manufacturing state tracing, and specifically relates to a device manufacturing state tracing method and system based on Internet of Things collection, which comprises the following steps: delimiting a manufacturing tracing boundary range according to unique part identification data, and determining a current manufacturing flow state in combination with production process behavior data. A state adaptation retrieval method is adopted to set a tracing retrieval mechanism in combination with the current manufacturing flow state, and a dynamic tracing adaptation mechanism is obtained through correction. According to working condition disturbance analysis data, a reliable manufacturing state portrait is constructed in cooperation with the dynamic tracing adaptation mechanism and the production process behavior data, and a dynamic quality judgment threshold is formulated. The reliable manufacturing state portrait is analyzed based on the dynamic quality judgment threshold, and a manufacturing state tracing strategy is generated; the present application can improve the accuracy of tracing range delimitation and the authenticity of manufacturing flow state recognition, and improve the accuracy of device manufacturing process anomaly diagnosis, quality tracing and process control.

Description

Technical Field

[0001] This invention belongs to the field of manufacturing status traceability technology, specifically a method and system for equipment manufacturing status traceability based on Internet of Things (IoT) data collection. Background Technology

[0002] With the widespread adoption of smart manufacturing and industrial IoT technologies, the equipment manufacturing industry has an increasingly urgent need for comprehensive, precise, and traceable control over the entire production process. Accurate manufacturing status traceability is the core foundation for ensuring product quality, locating production anomalies, and optimizing process parameters. Currently, most existing equipment manufacturing status traceability technologies employ fixed traceability logic, which still has many technical shortcomings that urgently need to be addressed in practical industrial applications.

[0003] First, existing traceability methods generally suffer from unclear traceability boundaries. Traditional solutions often rely on a single part code to define a fixed traceability range without dynamically adjusting the boundaries based on the production time and space scenario, process flow, equipment operating conditions, and abnormal states. This makes it easy for irrelevant and redundant data to be mixed in or for key production data to be missed, resulting in broken traceability data chains, chaotic range definition, and inability to adapt to complex production scenarios such as rework, cross-workstation transfer, and operating condition disturbances.

[0004] Secondly, the accuracy of manufacturing status identification is poor. Traditional technologies only determine basic states such as equipment start-up / shutdown and processing completion, without combining standard process routes with real-time process behavior data for benchmarking and matching. This makes it difficult to accurately distinguish between subdivided manufacturing states such as forward flow, rework, temporary storage, and cross-line transfer, and to identify hidden anomalies such as skipped steps or disordered sequences, resulting in low manufacturing process status reconstruction. Finally, traceability retrieval rules are severely rigid. Existing traceability retrieval mechanisms mostly use uniform and fixed retrieval logic, which cannot adapt to the differentiated scenarios of real-time part manufacturing flow status and workshop workstation partitioning. Retrieval trigger conditions, data filtering granularity, and result display formats cannot be dynamically adjusted, leading to problems such as poor retrieval targeting, low data matching accuracy, and low efficiency in anomaly tracing.

[0005] In summary, there is an urgent need for a method and system for tracing the manufacturing status of equipment based on IoT data collection to address the aforementioned shortcomings. Summary of the Invention

[0006] To overcome the shortcomings of existing technologies, this invention proposes a method and system for tracing the manufacturing status of equipment based on Internet of Things (IoT) data acquisition. This invention primarily addresses the problems of ambiguous traceability boundaries, inaccurate manufacturing status identification, and rigid traceability retrieval rules in existing status tracing methods.

[0007] The technical solution adopted by this invention to solve its technical problem is: a device manufacturing status traceability method based on IoT data acquisition provided by this invention, comprising: Collect equipment operating condition data, component unique identification data, and production process behavior data during the equipment manufacturing process. Determine the manufacturing traceability boundary based on the component unique identification data and determine the current manufacturing flow status by combining the production process behavior data.

[0008] A state-adaptive retrieval method is adopted to set a traceability retrieval mechanism in combination with the current manufacturing flow status. Based on the preset temporal grid algorithm, the traceability retrieval mechanism is modified according to the production workstation partition scenario to obtain a dynamic traceability adaptation mechanism.

[0009] Based on the degree of interference of equipment operating condition data on the steady-state parameters of the manufacturing process, operating condition disturbance analysis data is generated. A reliable manufacturing status profile is constructed by coordinating dynamic traceability and adaptation mechanisms and production process behavior data. Dynamic quality judgment thresholds are formulated by combining the evolution law of component wear with operating condition disturbance analysis data.

[0010] Based on dynamic quality judgment thresholds, a linked traceability and matching analysis is performed on the reliable manufacturing status profile to identify abnormal fluctuation sources and disturbed processes, thereby generating a manufacturing status traceability strategy.

[0011] The present invention provides a method for tracing the manufacturing status of equipment based on IoT data acquisition, wherein the steps for defining the manufacturing traceability boundary range include: The fields are broken down according to the preset coding rules, and the basic coding information in the unique identification data of the parts is extracted. The traceability query requirements are determined based on the parts currently being queried.

[0012] Based on the traceability query requirements, hierarchical traceability boundaries are defined, the spatiotemporal information of the current component identifier is associated, spatial boundaries are delineated, and the production timeline corresponding to the identifier is matched to define the time boundaries.

[0013] Matching identifiers bind processes, defining process flow boundaries, associating identifier-bound equipment operating condition data to determine data boundaries, and dynamically adjusting hierarchical traceability boundaries, spatial boundaries, temporal boundaries, process flow boundaries, and data boundaries based on different abnormal operating conditions to obtain the manufacturing traceability boundary range.

[0014] This invention provides a method for tracing the manufacturing status of equipment based on IoT data acquisition. The steps for determining the current manufacturing process status include: Based on the basic coding information, an ordered process node table is obtained by associating the corresponding product process route and bill of materials, and a standard process node table is formed by combining the manufacturing traceability boundary range.

[0015] Using the current component identifier as the primary key, extract the corresponding component's behavior events from the process behavior data and arrange them in ascending order by timestamp to form an event chain.

[0016] The event chain is aligned and matched with the standard process node table one by one, and the status is classified to obtain the process node where the component is currently located.

[0017] The current manufacturing flow status is obtained by identifying the flow direction and flow status of process nodes from forward flow, rework, temporary storage and waiting, and cross-line transfer.

[0018] This invention provides a method for tracing the manufacturing status of equipment based on IoT data acquisition. The steps for setting up a traceability retrieval mechanism include: Define search requirements corresponding to different manufacturing flow states, use the current component identifier as the search primary key, and build a search index system in combination with the manufacturing traceability boundary range.

[0019] For each type of manufacturing flow status, set corresponding search start thresholds and trigger rules, and set multiple search modes according to the differences in flow status.

[0020] Once the retrieval index system is activated, data is filtered according to three levels: manufacturing traceability boundary range, process stage, and data type.

[0021] Based on the current manufacturing flow status, match the types of data that need to be retrieved, set multiple display levels according to the query depth requirements corresponding to the flow status, and reset the search request and re-execute the search when the manufacturing flow status changes.

[0022] The present invention provides a method for tracing the manufacturing status of equipment based on IoT data acquisition, the steps of which to obtain a dynamic traceability adaptation mechanism include: According to the preset unified time step, the entire production cycle of the parts is divided into a time-series grid. The physical space of the workshop is divided into corresponding spatial grids using a preset minimum spatial unit. The time-series grid and the spatial grid are combined to form a spatiotemporal grid node. Combined with the production process, a preset flow route is generated.

[0023] The workshop is divided into functional zones according to its operational functions, which are then bound to a spatial grid, and production characteristics are recorded as a retrieval basis.

[0024] The traceability retrieval mechanism is mapped onto spatiotemporal grid nodes. Based on the actual process behavior records of parts, the spatiotemporal grid nodes that the parts pass through are matched to obtain the actual flow route.

[0025] The trajectory deviation between the actual flow route and the preset flow route is calculated from three aspects: time deviation, spatial deviation, and state deviation, and the correction link is determined.

[0026] Based on the search criteria and correction process, the search initiation threshold and triggering rules are modified, the search data filtering scope is modified by relying on spatial grids, and the data retrieval granularity rules are modified by relying on temporal grids.

[0027] Based on the partitioned scenarios and grid status, the logic for switching between cross-scenario searches was revised, and the display format of search results was revised based on the requirements of partitioned tasks.

[0028] The present invention provides a method for tracing the manufacturing status of equipment based on IoT data acquisition, the steps of which include generating operational disturbance analysis data: Establish a manufacturing steady-state benchmark parameter system for different processes, and select effective operating condition data within the manufacturing traceability boundary from equipment operating condition data.

[0029] The scenario where the actual operating parameters of the equipment deviate from the manufacturing steady-state reference parameter system is defined as operating condition disturbance, and different disturbance degree levels and disturbance forms are divided to obtain disturbance judgment classification rules.

[0030] For each data point in the effective operating condition data, the interference index is calculated by comparing them one by one based on the manufacturing steady-state benchmark parameter system, and the weights are set according to the process characteristics to calculate the interference score.

[0031] The disturbance scores of all valid operating condition data are weighted and summarized according to the weight ratio to obtain the current disturbance coefficient. The location, triggering cause and disturbance type of the disturbance are located to obtain the operating condition disturbance analysis data.

[0032] This invention provides a method for tracing the manufacturing status of equipment based on IoT data acquisition. The steps for constructing a reliable manufacturing status profile include: Using the unified timestamp of the time-series grid as a benchmark, the occurrence time of process behavior, the start and end time of operating condition disturbance, and the data acquisition time are calibrated and aligned, and then attached to the current component identifier to establish the correspondence between identifier, process behavior, operating condition disturbance, and grid partition.

[0033] Based on the needs of production control and traceability verification, we determine the dimensions of the profile and the corresponding data content to be filled in, and build a profile framework.

[0034] Based on the dynamic traceability and adaptation mechanism, matching data is retrieved according to the profile dimensions, and details of the operation flow status are extracted based on the production process behavior data. The operating condition disturbance analysis data is superimposed to supplement the stable operating status of the equipment.

[0035] By combining spatiotemporal grid nodes with retrieval criteria, the manufacturing trajectory status is improved, and behavioral disturbances and spatiotemporal status verification are performed.

[0036] For the portrait depiction dimensions, quantitative evaluation indicators are set, and the manufacturing status is characterized by a combination of qualitative description and quantitative values. The overall status level is divided according to the disturbance level and the abnormal status of the process.

[0037] Content is filled in sequentially according to the profile framework and bound to the dynamic traceability and adaptation mechanism to dynamically update the profile and obtain a reliable manufacturing status profile.

[0038] This invention provides a method for tracing the manufacturing status of equipment based on IoT data acquisition. The steps for setting dynamic quality judgment thresholds include: Collect component loss data, break down the process into different evolution stages according to processing sequence and cumulative loss level, determine the loss characteristics and inherent loss baseline of each stage, and the weight of the impact of production process on component loss.

[0039] Based on the operating condition disturbance analysis data, the impact of disturbances on losses is classified and matched. For each type of loss and its evolution stage, the disturbance influence coefficient is calculated, and a disturbance loss correlation library is constructed.

[0040] Based on the technical requirements and inspection specifications of the components, the judgment indicators are determined, and the basic quality judgment thresholds are set according to the inherent loss baseline and drawing standards to differentiate the judgment criteria at different levels.

[0041] Identify the current evolution stage, adjust the basic quality judgment threshold based on loss characteristics and influence weights, extract the disturbance influence coefficient corresponding to the current disturbance level from the disturbance loss association library, perform secondary correction, and generate a temporary dynamic threshold.

[0042] The temporary dynamic threshold is divided into threshold effective periods, and switching trigger conditions are set based on evolution stage jumps, disturbance level changes, process switching and processing batch replacement.

[0043] This invention provides a method for tracing the manufacturing status of equipment based on IoT data acquisition. The steps for generating a manufacturing status tracing strategy include: Extract measured data from each process period in the manufacturing status profile, compare each data point with the current dynamic quality judgment threshold, classify and label the quality anomaly types, and sort all anomaly nodes in chronological order to form anomaly time intervals.

[0044] Align the abnormal time interval with the production process behavior data along the time axis to locate the disturbed process corresponding to the quality anomaly type.

[0045] For the disturbed process, the abnormal fluctuation source is located by combining the operating condition disturbance analysis data, linking the spatiotemporal grid nodes and workstation partition information, investigating spatial flow-related fluctuation sources, and tracing the source of the body-related fluctuation source by combining the evolution stage of the component.

[0046] Based on process behavior, operating condition disturbance, spatiotemporal location, and component wear, multiple fluctuation sources are cross-validated, and the fluctuation sources that pass the validation are classified and labeled to obtain the fluctuation source type.

[0047] Based on the type of fluctuation source and the number of disturbed processes, traceability levels are divided, traceability directions for different scenarios are determined, and a manufacturing status traceability strategy is formulated for disturbed processes and fluctuation sources in combination with a dynamic traceability adaptation mechanism.

[0048] This invention provides a device manufacturing status traceability system based on IoT data acquisition, comprising: The full-domain data acquisition and sensing module is used to collect equipment operating condition data, component unique identification data, and production process behavior data during the equipment manufacturing process. Based on the component unique identification data, the manufacturing traceability boundary range is defined, and combined with the production process behavior data, the current manufacturing flow status is determined.

[0049] The retrieval mechanism adaptation module is used to set the traceability retrieval mechanism by combining the state adaptation retrieval method with the current manufacturing flow status. Based on the preset time-series grid algorithm, the traceability retrieval mechanism is modified according to the production workstation partition scenario to obtain a dynamic traceability adaptation mechanism.

[0050] The disturbance analysis modeling module is used to generate operating condition disturbance analysis data based on the degree of interference of equipment operating condition data on the steady-state parameters of the manufacturing process. It works in conjunction with the dynamic traceability adaptation mechanism and production process behavior data to build a reliable manufacturing state profile. It also combines the component wear evolution law with the operating condition disturbance analysis data to formulate dynamic quality judgment thresholds.

[0051] The linkage traceability matching module is used to perform linkage traceability matching analysis on the reliable manufacturing status profile based on the dynamic quality judgment threshold, identify the source of abnormal fluctuations and the disturbed process, and generate a manufacturing status traceability strategy.

[0052] The beneficial effects of this invention are as follows: 1. This invention, through the linkage of unique identifier data with time, space, process, and data boundaries, can accurately determine the manufacturing traceability scope, reduce interference from irrelevant data, and improve traceability efficiency and the accuracy of traceability scope delineation. By aligning the event chain with the standard process node table, the actual process and flow status of parts can be accurately identified, improving the authenticity of manufacturing flow status identification. Through the state-adaptive retrieval method and spatiotemporal grid correction mechanism, the retrieval rules can be dynamically adjusted according to different manufacturing states, workstation partitions, and actual flow deviations, improving the dynamic adaptability of the traceability retrieval mechanism.

[0053] 2. By constructing a reliable manufacturing status profile, process behavior, operational disturbances, spatiotemporal grids, and traceability mechanisms can be uniformly integrated, improving the reliability of the manufacturing status representation. By combining component wear evolution patterns and operational disturbance analysis data to formulate dynamic quality judgment thresholds, quality judgment standards can be adjusted according to changes in manufacturing status, reducing quality misjudgments caused by static thresholds and improving the accuracy of quality judgments. Through the linkage and matching of dynamic quality judgment thresholds with the reliable manufacturing status profile, abnormal fluctuation sources and disturbed processes can be accurately identified, and corresponding traceability strategies can be formulated based on the fluctuation source type, improving the efficiency of anomaly diagnosis. Attached Figure Description

[0054] The invention will now be further described with reference to the accompanying drawings.

[0055] Figure 1 This is a flowchart illustrating a device manufacturing status traceability method based on IoT data acquisition, provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of the process for obtaining the current manufacturing flow status in a device manufacturing status traceability method based on Internet of Things data acquisition provided in an embodiment of the present invention; Figure 3 This is a flowchart illustrating the dynamic traceability adaptation mechanism obtained in a device manufacturing status traceability method based on IoT data acquisition provided in an embodiment of the present invention. Figure 4 This is a schematic diagram of a device manufacturing status traceability system based on IoT data acquisition provided in an embodiment of the present invention. Detailed Implementation

[0056] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0057] like Figures 1 to 4 As shown in the figure, an embodiment of the present invention provides a method for tracing the manufacturing status of equipment based on IoT data acquisition, comprising: Collect equipment operating condition data, component unique identification data, and production process behavior data during the equipment manufacturing process. Determine the manufacturing traceability boundary based on the component unique identification data and determine the current manufacturing flow status by combining the production process behavior data.

[0058] The steps for defining the manufacturing traceability boundary include: The fields are broken down according to the preset coding rules, and the basic coding information in the unique identification data of the parts is extracted. The traceability query requirements are determined based on the parts currently being queried.

[0059] The unique identifier of the component is retrieved, and its fields are broken down according to the pre-set coding rules to extract the basic information contained in the code, including the component number, material batch, product model, hierarchical relationship, batch code, and production line. Simultaneously, it is matched with the background BOM list to determine the component's hierarchical assembly relationship within the product structure and clarify its subordinate level.

[0060] Using the unique identifier of the currently queried component as the core benchmark, traceability query requirements are differentiated: for single-item traceability, the individual component is used as the benchmark; for component traceability, the overall component identifier is used; and for whole-machine traceability, the unique code of the entire machine is locked, thereby determining the core reference target for boundary delineation.

[0061] Based on the traceability query requirements, hierarchical traceability boundaries are defined, the spatiotemporal information of the current component identifier is associated, spatial boundaries are delineated, and the production timeline corresponding to the identifier is matched to define the time boundaries.

[0062] Matching identifiers bind processes, defining process flow boundaries, associating identifier-bound equipment operating condition data to determine data boundaries, and dynamically adjusting hierarchical traceability boundaries, spatial boundaries, temporal boundaries, process flow boundaries, and data boundaries based on different abnormal operating conditions to obtain the manufacturing traceability boundary range.

[0063] The hierarchical traceability boundaries include: the smallest single-item boundary: only all associated data corresponding to the current single component identifier are collected, and only the processing, testing, transportation and assembly related content of the component itself is limited, without including information of other components, forming the smallest granularity traceability scope.

[0064] Component aggregation boundary: Based on the identification sub-field, retrieve the unique identifiers of all sub-components belonging to the same component, merge the production data corresponding to all associated identifiers, and define the traceability boundary of the entire component.

[0065] Whole machine full domain boundary: trace upwards along the coding level, summarize all parts and semi-finished product identifiers under the name of the whole machine, integrate all identifier binding data, and form the traceability boundary range of the whole equipment.

[0066] The system traces the physical locations of the component's entire journey, including production stations, machine tools, quality inspection stations, storage areas, and assembly lines. All production areas where the component's identifier appears are defined as spatial boundaries. Workshops, production lines, and equipment areas without a record of the component's movement are excluded, thus defining the scope of physical activity.

[0067] Extract the start time of the first scanning and online processing of the component's identification mark, as well as the end time of the finished product's production line exit, warehousing and sealing, and outbound circulation. Extract the start and end time intervals, including only all production records associated with the identification mark within this time period; exclude all historical and future irrelevant production data outside the interval.

[0068] The entire process of starting up, processing, stopping, transferring, quality inspection, rework, and assembly, all linked to the identification, is retrieved. Only the processes actually performed by the part are included in the traceability boundary; idle processes that were not touched or transferred, as well as processes for other products, are excluded, thus clarifying the complete process chain boundary.

[0069] Using the unique identifier of a component as an index, all equipment operating parameters, processing status, and equipment start-up and shutdown data attached to that identifier are filtered. Only the equipment operating condition information corresponding to the processing of that component is collected, and the equipment data generated by the processing of other components is separated to complete the definition of the operating condition data range.

[0070] When parts are reworked, disassembled and reassembled, batches are mixed, or materials are replaced, the component identification associations are updated simultaneously.

[0071] The identification data for rework, new processes, and replacement of parts are all incorporated into the original traceability boundaries. Independent parts are separated and new traceability boundaries are generated separately to ensure that the boundaries dynamically adapt to the production status.

[0072] The steps to determine the current manufacturing flow status include: Based on the basic coding information, an ordered process node table is obtained by associating the corresponding product process route and bill of materials, and a standard process node table is formed by combining the manufacturing traceability boundary range.

[0073] Standard process node table, such as: raw material feeding → rough machining → heat treatment → finish machining → inspection → assembly → warehousing, each node is marked with work station number, preset time, and dependencies.

[0074] Using the current component identifier as the primary key, extract the corresponding component's behavior events from the process behavior data and arrange them in ascending order by timestamp to form an event chain.

[0075] The event chain is aligned and matched with the standard process node table one by one, and the status is classified to obtain the process node where the component is currently located.

[0076] Alignment matching: Map each actual event to a standard process node to determine which process has been completed.

[0077] Status Classification: Each process node corresponds to a status value. Common classifications are as follows:

[0078] The current manufacturing flow status is obtained by identifying the flow direction and flow status of process nodes from forward flow, rework, temporary storage and waiting, and cross-line transfer.

[0079] Forward flow: proceeding sequentially according to the process route, which is a normal state.

[0080] Rework and Return: Behavioral data shows that parts flow back from downstream workstations to upstream workstations (e.g., if the finishing process fails, it is returned to the roughing process). The number of reworks and the reason for each rework need to be recorded.

[0081] Temporarily waiting: The current process has been completed, but the next process has not yet started, and the preset waiting time limit has been exceeded, indicating that there is a backlog or scheduling problem on the production line.

[0082] Cross-line transfer: The transfer of parts to another production line or outsourced unit requires confirmation based on scheduling system data.

[0083] Based on the above assessments, a structured description of the current manufacturing flow status is generated, such as: Part SN-20240531-0078, currently in the finishing process, status: in progress, time taken: 2.3 hours, preset time: 1.8 hours, timeout, this is the second flow, the next planned process is final inspection, and the estimated flow time window is: 14:00–14:30.

[0084] A state-adaptive retrieval method is adopted to set a traceability retrieval mechanism in combination with the current manufacturing flow status. Based on the preset temporal grid algorithm, the traceability retrieval mechanism is modified according to the production workstation partition scenario to obtain a dynamic traceability adaptation mechanism.

[0085] The steps to set up a traceability retrieval mechanism include: Define search requirements corresponding to different manufacturing flow states, use the current component identifier as the search primary key, and build a search index system in combination with the manufacturing traceability boundary range.

[0086] Search requests may include: normal processing / assembly / transfer status: requests mainly focus on viewing real-time working conditions, process progress, and subsequent flow planning.

[0087] Quality inspection status assessment: The main requests are to retrieve historical processing parameters, test data, and reference records of similar parts.

[0088] Repair and rectification status: The requests mainly involve tracing back the process in which the defect occurred, checking abnormal equipment parameters, and verifying operation records.

[0089] Process paused / delayed status: The main requests are to find the cause of the pause and to verify the connection information between the preceding and following processes.

[0090] Abnormal status of process jump / sequence disorder: The main requests are to restore the actual flow trajectory and compare the differences of standard processes.

[0091] Finished product completion and warehousing status: The main requirement is to summarize the entire process archives and generate complete traceability certificates.

[0092] Set the part identification code as the primary index, and all search operations will revolve around this code.

[0093] Set the start and end times, production plant area, and processing line of the traceability boundary as secondary range indexes.

[0094] Set the current process, associated processing equipment, and operator number as a third-level detailed index.

[0095] Establish an index-based lookup table so that the identification code can quickly retrieve all process, working condition, and material data within the boundary.

[0096] For each type of manufacturing flow status, set corresponding search start thresholds and trigger rules, and set multiple search modes according to the differences in flow status.

[0097] In real-time processing mode: Triggering instant retrieval, the retrieval channel remains open as long as the equipment continues to generate operating data and update process actions.

[0098] In transit / movement state: Trigger node retrieval, which is initiated and refreshed only when the part leaves the current workstation and enters a new station.

[0099] In quality inspection status: Trigger backtracking retrieval to automatically retrieve all previous processing steps and parameter records for the part.

[0100] In the rework / rectification state: triggers anomaly-oriented search, prioritizing the search for defect locations, abnormal parameters, and non-conformance judgment records.

[0101] When a process is stalled: trigger a breakpoint search to focus on the equipment status and process handover records before and after the stall.

[0102] When the product is in the finished product stage: trigger a full archive retrieval, and retrieve the entire set of production data within the boundary at once.

[0103] Based on differences in workflow status, multiple switchable search modes are set to meet different usage scenarios such as daily viewing, problem investigation, and batch verification, including: Single-point real-time retrieval mode: Adapts to processing and transfer status, and only retrieves the latest process and equipment operation data at the current moment.

[0104] Forward process retrieval mode: Adapted to progress verification scenarios, it retrieves process records sequentially from launch to the current node according to the order of the work processes.

[0105] Reverse backtracking retrieval mode: Adapted to fault and defect troubleshooting scenarios, it allows you to trace back all previous production data from the current abnormal node.

[0106] Linked search mode: Adapts to the assembly status and retrieves traceability information of other components that are assembled with the current part.

[0107] Batch retrieval mode: When a batch of parts are in the same circulation state, retrieve the traceability data of all parts in the same batch in batch.

[0108] Once the retrieval index system is activated, data is filtered according to three levels: manufacturing traceability boundary range, process stage, and data type.

[0109] The first layer: the boundary range screening is strictly limited to the manufacturing traceability boundary defined in the early stage, excluding irrelevant data from other workshops, other batches, and other time periods.

[0110] The second layer: Process filtering based on the current process node, only the data of the process and its related preceding and following processes are filtered, and idle process records that are not involved are skipped.

[0111] The third layer: Business data type filtering focuses on processing status (emphasizing equipment operating parameters), quality inspection status (emphasizing test result data), and rework status (emphasizing abnormal alarm data).

[0112] Based on the current manufacturing flow status, match the types of data that need to be retrieved, set multiple display levels according to the query depth requirements corresponding to the flow status, and reset the search request and re-execute the search when the manufacturing flow status changes.

[0113] Processing status: Real-time operating data such as equipment speed, temperature, pressure, and processing feed are retrieved in conjunction with process start and stop data.

[0114] Transfer status: Link to retrieve workstation transfer records, loading and unloading operation data, and site location change data.

[0115] Quality inspection status: Linked retrieval of dimensions, performance test data, pass / fail results, and quality inspection personnel operation records.

[0116] Repair status: Linkage to retrieve historical abnormal alarms, defect descriptions, maintenance and debugging parameters, and secondary processing records.

[0117] Completion status: Linkage to retrieve all process behaviors, full process conditions, quality inspection reports, and material traceability data.

[0118] Three display levels—simple, standard, and advanced—are set up to intuitively present the search results.

[0119] Simple display: Shows the current location, process, workflow duration, and basic status conclusions, suitable for on-site operators to view quickly.

[0120] Standard display: Completely lists the process flow, key equipment parameters, and operation shift information to meet the needs of daily production control and verification.

[0121] In-depth display: Includes anomaly comparison data, process standard references, related parts information, and problem tracing clues for quality analysis and troubleshooting.

[0122] When a part switches from processing to quality inspection, or from normal circulation to rework, the system immediately recognizes the status change, clears the original search and filtering conditions, matches the corresponding search rules according to the new circulation status, re-filters and retrieves data, and synchronously refreshes the content displayed on the search page to ensure that the search results always match the real-time status.

[0123] The steps to obtain the dynamic traceability adaptation mechanism include: According to the preset unified time step, the entire production cycle of the parts is divided into a time-series grid. The physical space of the workshop is divided into corresponding spatial grids using a preset minimum spatial unit. The time-series grid and the spatial grid are combined to form a spatiotemporal grid node. Combined with the production process, a preset flow route is generated.

[0124] Based on the speed of production and processing and the time consumption of each process, a uniform time step is set, and the entire production cycle of the part is divided into continuous and equally spaced time sequence grid units. Each grid represents a production period of fixed duration, and the grid start time, end time, and time sequence number are marked.

[0125] The spatial dimension grid is divided into corresponding spatial grids with a single workstation or small work area as the smallest spatial unit. Each spatial grid is bound to a unique grid code, corresponding to a fixed workstation, processing equipment, or work area.

[0126] By combining temporal and spatial grids, spatiotemporal grid nodes are formed, with each node uniquely corresponding to a specific time period and workstation area. Simultaneously, component identifiers, process behaviors, and equipment operating condition data are attached to the corresponding grid nodes.

[0127] Based on standard production processes, the standard grid movement trajectory of parts for normal processing, transfer, quality inspection, and warehousing is preset in the grid model, and the temporal-spatial grid sequence that the parts should pass through in sequence is specified.

[0128] The workshop is divided into functional zones according to its operational functions, which are then bound to a spatial grid, and production characteristics are recorded as a retrieval basis.

[0129] The workshop is divided into six basic zones according to its operational functions: machining zone, assembly zone, quality inspection zone, rework zone, material storage zone, and finished product storage zone.

[0130] Each spatial grid is assigned and bound to its corresponding functional partition, clearly defining which workstation partition each grid belongs to. The production characteristics of each partition are recorded: processing data generation density, workstation flow rate, process dwell time, common anomaly types, and data collection frequency, serving as the basis for retrieval.

[0131] The traceability retrieval mechanism is mapped onto spatiotemporal grid nodes. Based on the actual process behavior records of parts, the spatiotemporal grid nodes that the parts pass through are matched to obtain the actual flow route.

[0132] The original primary key index, range filtering conditions, and triggering rules of the retrieval are mounted one by one to each spatiotemporal grid node and workstation partition.

[0133] Check the performance of the original search rules one by one in different partitions and grid time periods, and record the adaptation defects: such as fixed search frequency not matching high and low density data, data breakage in cross-partition search, search failure in grid stagnation state, and deviation of search results due to time sequence misalignment.

[0134] The trajectory deviation between the actual flow route and the preset flow route is calculated from three aspects: time deviation, spatial deviation, and state deviation, and the correction link is determined.

[0135] Timing deviation: The difference between the actual time of entering and leaving the grid and the standard time.

[0136] Spatial deviation: The deviation between the actual station grid and workstation zoning and the planned path.

[0137] Status deviation: The situation where the actual process status within the grid does not match the standard operation status.

[0138] Identify the retrieval failure scenarios corresponding to classification label bias and pinpoint the specific aspects of the existing mechanism that need to be corrected.

[0139] Based on the search criteria and correction process, the search initiation threshold and triggering rules are modified, the search data filtering scope is modified by relying on spatial grids, and the data retrieval granularity rules are modified by relying on temporal grids.

[0140] High-density processing partition grid: With frequent part processing actions and rapid refresh of working condition data, the retrieval and polling interval is shortened. As soon as new process data or equipment parameter changes occur within the grid, the retrieval and update are triggered immediately.

[0141] Quality inspection and rework zone grid: For processes with long dwell times, a node-based triggering method is used, and the search is only initiated when the quality inspection judgment is completed or the rework process is started or stopped.

[0142] Material temporary storage zoning grid: When there is no operation for a long time, the frequency of automatic retrieval is reduced, and the retrieval is only activated when manually queried or when the grid position changes.

[0143] Cross-partition boundary grid: Add a partition switching trigger signal. When a part leaves the current partition grid and enters a new partition, the search rule switching is automatically triggered.

[0144] Abandoning the original fixed traceability retrieval scope, the system uses the current spatial grid of the part as the core, finely filtering all related data within this grid. It also appropriately links data from adjacent upstream and downstream grids to ensure data continuity during short-distance transfers between workstations. Data from unrelated partitions and grids is strictly isolated to prevent redundant information from being mixed into the retrieval results. When a part moves across multiple grids within the same partition, the retrieval scope expands and contracts synchronously with the spatial grid. When transferring across partitions, the retrieval boundary switches synchronously to the new partition's grid system.

[0145] Short-time fine grid: for precision machining and rapid processes, it retrieves millisecond-level and second-level fine process parameters and process details.

[0146] Long-term coarse grid: corresponding to static storage and long-term assembly processes, it summarizes the overall status data within a time period and simplifies fragmented and repetitive parameters.

[0147] For nodes with timing deviations or grid time discontinuities, the algorithm automatically extends adjacent grid time periods to fill in missing data and correct data omissions caused by fixed-time truncation.

[0148] Based on the partitioned scenarios and grid status, the logic for switching between cross-scenario searches was revised, and the display format of search results was revised based on the requirements of partitioned tasks.

[0149] When a part remains in the processing area grid, the system automatically switches to real-time operating condition tracking and retrieval logic, focusing on capturing data related to equipment operation and cutting processes.

[0150] When a part enters the quality inspection area grid: the reverse backtracking retrieval logic is automatically switched to retrieve the historical production records of all previously processed grids.

[0151] When a part enters the rework area grid: the abnormal orientation search logic is automatically switched, prioritizing the matching of deviation grids, alarm data, and non-conforming records.

[0152] Parts flow to the warehouse area grid: Automatically switch to full archive retrieval logic to summarize the complete set of traceability data from all time-series grids and partition grids.

[0153] Processing station grid: Concisely displays real-time location, current parameters, and process progress, meeting the needs for quick on-site viewing.

[0154] Quality inspection and rework zone grid: Comparative display of standard process grid data and actual grid data, intuitively presenting deviation issues.

[0155] The whole machine assembly partition grid: link and retrieve the grid traceability information of all subordinate parts to display the whole assembly related traceability content.

[0156] Based on the degree of interference of equipment operating condition data on the steady-state parameters of the manufacturing process, operating condition disturbance analysis data is generated. A reliable manufacturing status profile is constructed by coordinating dynamic traceability and adaptation mechanisms and production process behavior data. Dynamic quality judgment thresholds are formulated by combining the evolution law of component wear with operating condition disturbance analysis data.

[0157] The steps for generating operating condition disturbance analysis data include: Establish a manufacturing steady-state benchmark parameter system for different processes, and select effective operating condition data within the manufacturing traceability boundary from equipment operating condition data.

[0158] Retrieve the official process documents for the current processing steps of the component, the equipment's factory rated parameters, and the optimal operating data of historical qualified products to determine the standard steady-state reference values ​​for various operating parameters.

[0159] Based on the allowable fluctuation range of the process, the upper and lower limits of the steady-state normal range of each parameter are defined, and a small normal fluctuation threshold is set to distinguish the boundary between normal fluctuation and abnormal disturbance.

[0160] The system is categorized and bound by equipment, workstation, and process, with each machine tool and processing stage matched with its own dedicated steady-state standard to avoid the mixing of reference parameters across processes.

[0161] The parameters such as temperature, pressure, rotational speed, vibration, load current, and feed rate are assigned weights based on their impact on quality. Core forming parameters have higher weights, while auxiliary cooling and auxiliary transmission parameters have lower weights.

[0162] Define the manufacturing traceability boundary and filter all data collected within the boundary based on the unique identifier of the part, the processing time range, and the corresponding workstation equipment number. Remove data that is outside the traceability period, does not belong to the current part processing, or is irrelevant data generated by other workstation equipment, retaining only the truly matching processing condition records. Organize the basic fields of the original data into a unified format including collection time, equipment number, parameter name, real-time monitoring value, sampling frequency, corresponding process number, and part code.

[0163] The scenario where the actual operating parameters of the equipment deviate from the manufacturing steady-state reference parameter system is defined as operating condition disturbance, and different disturbance degree levels and disturbance forms are divided to obtain disturbance judgment classification rules.

[0164] Disturbance definition: When the actual operating parameters of the equipment deviate from the preset steady-state normal range, it is considered a disturbance in the operating conditions. The greater the deviation and the longer the duration, the higher the degree of disturbance.

[0165] The disturbance is classified into four levels: slight disturbance, moderate disturbance, severe disturbance, and violent disturbance, each corresponding to a different percentage threshold for parameter deviation.

[0166] The criteria for distinguishing three types of disturbances are: instantaneous pulse disturbances, periodic repetitive disturbances, and slow drift disturbances.

[0167] For each data point in the effective operating condition data, the interference index is calculated by comparing them one by one based on the manufacturing steady-state benchmark parameter system, and the weights are set according to the process characteristics to calculate the interference score.

[0168] For each operating parameter, such as temperature, speed, pressure, and vibration, the steady-state reference value is compared one by one.

[0169] Calculate the interference index: Absolute deviation = actual operating value of equipment - steady-state baseline mean.

[0170] Relative deviation rate = absolute deviation ÷ steady-state baseline mean × 100%.

[0171] The statistical parameters include the duration of the deviation from the steady-state range, the number of fluctuations per unit time, and the maximum deviation peak value.

[0172] By combining the weights set according to the characteristics of the process, the interference score of a single parameter is calculated. The higher the score, the stronger the disruption to the manufacturing steady state.

[0173] The disturbance scores of all valid operating condition data are weighted and summarized according to the weight ratio to obtain the current disturbance coefficient. The location, triggering cause and disturbance type of the disturbance are located to obtain the operating condition disturbance analysis data.

[0174] The interference scores of all individual operating parameters are weighted and summed according to their respective weight ratios to obtain the overall comprehensive disturbance coefficient of the current equipment, representing the global interference level. The relationships between parameter changes are analyzed, such as vibration and current changes caused by speed fluctuations, to identify coupled disturbances formed by the mutual influence of parameters. It is determined whether superimposed disturbances amplify the deviation effect, distinguishing between two types of interference: single small disturbances and multiple superimposed strong disturbances.

[0175] Based on data timestamps, the start and end times of disturbances, as well as the duration of each disturbance, are precisely located, and the frequency of disturbances within the same process is statistically analyzed. The corresponding processing station, machine tool, and process node are linked to determine the physical location of the disturbance. Combined with process behavior and on-site production conditions, potential causes are investigated: equipment wear and aging, differences in raw material materials, manual modification of process parameters, changes in workshop temperature and humidity, and fluctuations in power grid voltage, etc., and the causes of disturbances are categorized and labeled.

[0176] Disturbances are categorized based on their parameter deviation characteristics: sudden instantaneous disturbances, periodic reciprocating disturbances, slow offset disturbances, and step-change disturbances. Manufacturing risks associated with disturbances are assessed based on process experience: pressure disturbances can easily cause dimensional inconsistencies in parts, vibration disturbances can easily produce surface finishing defects, and temperature disturbances can affect material forming stability. The current manufacturing workflow is compared to determine whether disturbances will hinder normal process progress, cause downtime, or produce defective products.

[0177] The steps to build a trustworthy manufacturing state profile include: Using the unified timestamp of the time-series grid as a benchmark, the occurrence time of process behavior, the start and end time of operating condition disturbance, and the data acquisition time are calibrated and aligned, and then attached to the current component identifier to establish the correspondence between identifier, process behavior, operating condition disturbance, and grid partition.

[0178] Using the unified timestamp of the time-series grid as a benchmark, the occurrence time of process behavior, the start and end time of operating condition disturbance, and the data acquisition time are all calibrated and aligned to eliminate clock deviation and ensure that the behavior and equipment disturbance status within the same time period correspond one-to-one.

[0179] By combining workstation zoning and spatial grid coding, the workstations where processes occur, the points where disturbances occur, and the actual areas where parts stay are matched together to clarify the physical operation location corresponding to each piece of data.

[0180] Based on the needs of production control and traceability verification, we determine the dimensions of the profile and the corresponding data content to be filled in, and build a profile framework.

[0181] Basic identity dimensions: part code, product, material batch, processing equipment, and production line station.

[0182] Workflow behavior dimensions: work sequence, work actions, workflow duration, shift personnel, and work compliance status.

[0183] Equipment operation disturbance dimensions: steady-state deviation of each parameter, disturbance level, disturbance frequency, and operational stability.

[0184] Spatiotemporal trajectory dimension: time-series grid travel route, workstation zone flow path, progress speed deviation.

[0185] Quality risk dimensions: potential processing defects caused by disturbances, and quality uncertainties caused by process abnormalities.

[0186] Comprehensive compliance status dimensions: whether the overall process meets the standards, whether the equipment operating conditions meet the standards, and whether the process flow is standardized.

[0187] Based on the dynamic traceability and adaptation mechanism, matching data is retrieved according to the profile dimensions, and details of the operation flow status are extracted based on the production process behavior data. The operating condition disturbance analysis data is superimposed to supplement the stable operating status of the equipment.

[0188] Based on the current manufacturing flow status of the part, the corresponding retrieval mode of the adaptation mechanism is invoked: real-time grid data is retrieved for normal processing, historical data is retrieved for abnormal rework, and full-cycle summary data is retrieved for the completed status.

[0189] Based on the temporal grid and workstation partition boundaries, the system automatically partitions and retrieves process behavior records and working condition disturbance analysis results within the corresponding area, and synchronously switches the retrieval range when transferring between partitions.

[0190] The entire process is organized chronologically to reconstruct the complete flow of parts from material loading, processing, transfer, quality inspection to completion, and to record the start and end points and actual time consumption of each process.

[0191] Determine the compliance of process execution and distinguish between different behavioral states such as standard process operation, skipping steps in the process, repeated rework, long-term downtime, and disordered process sequence.

[0192] Record human and equipment operation behaviors such as operator actions, equipment start-up and shutdown operations, and material changes, improve the detailed description of the operation process, and form a status basis at the behavioral level.

[0193] By matching and binding the disturbance data corresponding to each process node to the process behavior node, the fluctuation and deviation of equipment parameters during each operation can be clearly identified.

[0194] Based on the disturbance level and the comprehensive disturbance coefficient, the operational stability of a single process and the overall processing is marked, and the states of steady-state operation, slight disturbance, severe disturbance, and violent abnormal disturbance are distinguished.

[0195] Based on the type and scope of disturbance, indicate the negative impact of abnormal parameters such as vibration, temperature, pressure, and speed on the current process, and supplement the status characteristics at the equipment operation level.

[0196] By combining spatiotemporal grid nodes with retrieval criteria, the manufacturing trajectory status is improved, and behavioral disturbances and spatiotemporal status verification are performed.

[0197] By comparing the actual spatiotemporal grid path traversed by the part with the preset standard flow grid, it is determined whether the trajectory deviates from the planned route.

[0198] The duration and frequency of parts entering and leaving different zones such as machining area, quality inspection area, rework area, and temporary storage area are statistically analyzed to characterize the spatial distribution.

[0199] By comparing the time taken for standard processes with the actual processing time, the timing of production progress—whether it is ahead of schedule, normal, or lagging behind—can be determined, thus enriching the spatiotemporal dimension profile.

[0200] Mutual verification of behavior and disturbance: During periods when process delays or rework occur, check whether there are simultaneous severe disturbances in the operating conditions, confirm that the abnormal states match each other, and eliminate false data records.

[0201] Verification between time, space, and state: The grid partition where the part is located must have a corresponding process type and equipment disturbance characteristics that conform to the production characteristics of that area. If they do not conform, the data must be checked and corrected.

[0202] For the portrait depiction dimensions, quantitative evaluation indicators are set, and the manufacturing status is characterized by a combination of qualitative description and quantitative values. The overall status level is divided according to the disturbance level and the abnormal status of the process.

[0203] Quantitative evaluation indicators may include: process compliance completion rate, overall comprehensive disturbance coefficient, workflow deviation rate, frequency of abnormal disturbances, and percentage of time spent in a particular area.

[0204] Combining qualitative descriptions with quantitative data, the manufacturing status is depicted in two ways: it indicates whether the process is currently in the processing, quality inspection, or rework stage, and it also uses data to reflect the stability and compliance level.

[0205] Based on the level of disturbance and the abnormality of the process, the overall status is divided into four levels: excellent steady state, normal small fluctuation, slight abnormality, and high-risk abnormality.

[0206] Content is filled in sequentially according to the profile framework and bound to the dynamic traceability and adaptation mechanism to dynamically update the profile and obtain a reliable manufacturing status profile.

[0207] This process summarizes all information, including identity information, process behavior, equipment disturbances, spatiotemporal trajectories, risk levels, and overall status, and fills in the content sequentially according to a pre-defined profile framework. It analyzes status change trends, marking transition points from steady state to disturbance, and from normal flow to abnormal stagnation, clearly presenting the state evolution pattern of the entire manufacturing process. It also summarizes current problem locations, potential quality hazards, and equipment operation risks, separately marking risk warnings within the profile.

[0208] Establish a linkage interface between the profiling and dynamic traceability adaptation mechanism. When new process behaviors are generated or new operational disturbance data is detected, data updates are automatically triggered. When parts undergo workstation switching, process status changes, or disturbance levels change, the corresponding flow status, operational stability, and risk level content in the profiling are simultaneously corrected. This ensures that every change in the manufacturing process is reflected in the status profiling in real time, always closely matching the actual production situation on site.

[0209] The steps for determining dynamic quality judgment thresholds include: Collect component loss data, break down the process into different evolution stages according to processing sequence and cumulative loss level, determine the loss characteristics and inherent loss baseline of each stage, and the weight of the impact of production process on component loss.

[0210] The number of component losses can include the component's material properties, processing technology, structural characteristics, design life, and historical scrap / defective product ledger. It can also summarize the loss forms of the component throughout the entire manufacturing process from raw materials, processing and forming, assembly to completion, including types such as wear, deformation, stress fatigue, surface damage, and dimensional deviation.

[0211] Based on the processing sequence and the degree of accumulated losses, different evolutionary stages are identified, and the loss characteristics of each stage are clearly defined: Initial break-in phase: Parts / processing tools wear out slowly, and there are minimal fluctuations in size and performance, which is the initial low-wear state.

[0212] Stable loss stage: Losses accumulate at a uniform rate, and parameter changes are stable; this is the mainstream stage of production.

[0213] Accelerated wear stage: The wear rate increases significantly, and the shape and performance gradually deviate from the design standards.

[0214] Critical failure stage: The loss is close to the acceptable limit, and defective products are very likely to be produced. It is close to the scrap / rework boundary.

[0215] By statistically analyzing the normal wear and parameter offset range of the same type of parts under no significant operating conditions, and considering the changes in processing time and quantity at each stage, a natural wear baseline curve is formed, and the corresponding normal parameter fluctuation range for each stage is recorded.

[0216] Clearly define the impact weight of each production process on component wear. For example, the cutting process is prone to surface wear, and the stamping process is prone to deformation wear. Mark high-wear processes and low-wear processes, and distinguish the differences in wear evolution under different processes.

[0217] Based on the operating condition disturbance analysis data, the impact of disturbances on losses is classified and matched. For each type of loss and its evolution stage, the disturbance influence coefficient is calculated, and a disturbance loss correlation library is constructed.

[0218] We categorize and statistically analyze the impact of disturbances by type and level, and how different operating conditions affect the original wear rate of components: minor disturbances: slightly increase the wear rate, and slightly increase the parameter offset.

[0219] Normal / severe disturbances: significantly accelerate loss accumulation, shorten the duration of the stable loss phase, and enter the accelerated loss range earlier.

[0220] Severe disturbances: instantaneously cause local deformation, cracks, and dimensional deviations, directly leading to sudden quality defects.

[0221] For each type of loss and each stage of loss, calculate the disturbance impact coefficient: using the natural loss under no disturbance as a benchmark, calculate the loss increment and parameter offset amplification factor caused by different levels of disturbance.

[0222] The system establishes a database that maps "loss stage + process type + disturbance level + loss increment + parameter offset magnitude" to form a database of the linkage between disturbance and loss.

[0223] Based on the technical requirements and inspection specifications of the components, the judgment indicators are determined, and the basic quality judgment thresholds are set according to the inherent loss baseline and drawing standards to differentiate the judgment criteria at different levels.

[0224] Based on the technical requirements and inspection specifications of the parts, determine the judgment indicators, which may include dimensional, geometric tolerance, surface quality, performance, etc., and clarify the upper and lower limits of static acceptance for each indicator before leaving the factory.

[0225] During the initial break-in period of the components, under conditions of no operational disturbance, the basic quality judgment threshold is set with reference to the natural wear baseline and in conjunction with the drawing standards.

[0226] The judgment criteria are divided into two levels: warning threshold: when the loss is close to the upper limit of normal loss, it needs to be closely monitored.

[0227] Non-conformance threshold: Products meeting quality defect standards are judged as defective. Both types of thresholds require dynamic adjustment in the future.

[0228] Identify the current evolution stage, adjust the basic quality judgment threshold based on loss characteristics and influence weights, extract the disturbance influence coefficient corresponding to the current disturbance level from the disturbance loss association library, perform secondary correction, and generate a temporary dynamic threshold.

[0229] Based on the quantity processed, cumulative processing time, historical wear records, and wear evolution curves, determine whether the component is currently in the break-in, stabilization, acceleration, or critical failure stage.

[0230] Initial break-in phase: The increase in wear and tear is minimal. The initial baseline threshold is used, and the judgment criteria are tightened to ensure the initial quality of the product.

[0231] Stable loss stage: Losses accumulate at a uniform rate, and the threshold is slightly relaxed according to the normal loss amount of the stage to match the normal process fluctuation.

[0232] Accelerated Loss Phase: Losses continue to increase, the allowable range of parameter deviation is adjusted upwards simultaneously, and the warning threshold is tightened to provide early risk alerts.

[0233] Critical failure stage: Strictly adhere to the factory's maximum hard threshold, significantly tighten the criteria for judging non-conformity, and prevent defective products from being released.

[0234] The calculation is performed by adding a disturbance correction coefficient to the base threshold after the phase correction: if there is a slight disturbance: the warning threshold is slightly relaxed, and the non-compliance threshold remains unchanged.

[0235] If there is a general / severe disturbance: simultaneously relax the acceptable range, trigger the early warning threshold in advance, and increase the detection frequency.

[0236] Severe disturbance exists: Temporarily tighten the non-compliance threshold, and once the parameter exceeds the standard, it will be directly judged as defective, and the source of the disturbance will be marked.

[0237] The temporary dynamic threshold is divided into threshold effective periods, and switching trigger conditions are set based on evolution stage jumps, disturbance level changes, process switching and processing batch replacement.

[0238] Set short-term dynamic thresholds: For instantaneous disturbances and short-term processes, the threshold takes effect only once. Once the disturbance disappears or the process ends, it automatically reverts to the base threshold for the current loss stage.

[0239] Set medium- to long-term dynamic thresholds: For continuous disturbances and components entering a new wear stage, the thresholds remain in effect until the wear stage switches or the operating conditions return to steady state.

[0240] Configure automatic switching trigger conditions: specify the signals that trigger threshold updates: loss stage jump, disturbance level change, process switch, processing batch change. If any condition is met, the quality judgment threshold will be automatically recalculated and updated.

[0241] Based on dynamic quality judgment thresholds, a linked traceability and matching analysis is performed on the reliable manufacturing status profile to identify abnormal fluctuation sources and disturbed processes, thereby generating a manufacturing status traceability strategy.

[0242] The steps for generating a manufacturing status traceability strategy include: Extract measured data from each process period in the manufacturing status profile, compare each data point with the current dynamic quality judgment threshold, classify and label the quality anomaly types, and sort all anomaly nodes in chronological order to form anomaly time intervals.

[0243] Extract the actual quality indicators, equipment operating parameters, and measured data of components for each process and time period from the manufacturing status profile, and compare them one by one with the currently effective dynamic judgment threshold.

[0244] Classify and label anomaly types: If the node is close to the warning threshold, it is marked as a potential anomaly node and should be given special attention.

[0245] Nodes exceeding the warning threshold or failing to meet the unqualified threshold are marked as general abnormal nodes.

[0246] Nodes exceeding the unqualified threshold or failing to break the limit hard threshold are marked as severely abnormal nodes.

[0247] Exceeding the factory-set hard threshold: Marked as a fatal defect node.

[0248] By combining the time-series information in the trusted manufacturing status profile, all abnormal nodes are sorted out in chronological order, and the start and end times, duration, and deviation of indicators are recorded to form abnormal time intervals.

[0249] Align the abnormal time interval with the production process behavior data along the time axis to locate the disturbed process corresponding to the quality anomaly type.

[0250] Distinguish the scope of the abnormal impact and classify the affected processes into three categories: Directly affected processes: Abnormal indicators and operating condition disturbances occur simultaneously with the process, and the process operation is directly disturbed.

[0251] Indirectly disturbed processes: Abnormalities generated in previous processes are propagated to subsequent processes, causing passive fluctuations in the parameters and product status of this process.

[0252] Chain-disrupted processes: A single point of abnormality triggers batch fluctuations in multiple consecutive processes, forming a process chain disturbance.

[0253] The frequency, amplitude, and duration of abnormalities for each affected process are statistically analyzed, and the degree of impact is ranked to determine the core affected processes and the secondary affected processes.

[0254] For the disturbed process, the abnormal fluctuation source is located by combining the operating condition disturbance analysis data, linking the spatiotemporal grid nodes and workstation partition information, investigating spatial flow-related fluctuation sources, and tracing the source of the body-related fluctuation source by combining the evolution stage of the component.

[0255] For the identified disturbed processes, retrieve the corresponding time period and equipment disturbance analysis data, and match the disturbance level, disturbance type, comprehensive disturbance coefficient, and parameter deviation.

[0256] Breaking down the sources of fluctuation layer by layer: Single parameter fluctuation source: If only one parameter such as speed, temperature, pressure, vibration, or current continuously deviates from the steady state, it is determined to be a single equipment unit failure, sensor drift, or single-channel load abnormality.

[0257] Multi-parameter coupled fluctuation source: Synchronous deviation of multiple operating parameters is identified as equipment instability, transmission system failure, power grid fluctuation, and workshop environmental interference.

[0258] Instantaneous pulse fluctuation source: A sudden jump in parameters followed by rapid recovery over a short period of time is identified as an instantaneous impact, electromagnetic interference, brief power outage, or contact with foreign objects.

[0259] Slow drift fluctuation source: The parameters gradually and continuously shift. Combined with the evolution law of component wear, it is determined to be tool or mold wear, equipment component aging, and accelerated natural wear of components.

[0260] By comparing the spatiotemporal grid trajectory and workstation partition information in the trusted manufacturing status profile, the physical location corresponding to the abnormal node is checked: if the abnormality only occurs in a single spatial grid / single workstation, it is determined to be a single-point workstation-specific fluctuation source, and the problem is concentrated in this equipment and this area environment.

[0261] Anomalies occurring across multiple grids or multiple times within the same functional zone are identified as common fluctuation sources within the zone, such as zone power supply, centralized gas supply, overall temperature and humidity, or regional electromagnetic interference.

[0262] Anomalies are transferred synchronously to different workstations as the parts are moved: they are determined to be sources of fluctuations in the parts themselves, with the root cause being defects in the preceding materials or the semi-finished products themselves.

[0263] By comparing the actual flow grid of the parts with the standard preset grid, if there are violations such as skipping grids, route deviation, or cross-area delays, the abnormal flow of the process is determined to be a secondary source of fluctuation.

[0264] Take the current wear stage of the component and its inherent wear baseline, and perform correlation analysis with abnormal indicators to determine the source of the fluctuation. The component is in the accelerated / critical failure stage, and the wear rate is far beyond the baseline: it is determined that the abnormal condition is caused by the component itself wearing out too quickly.

[0265] If multiple parts from the same batch exhibit similar fluctuations during the same process, it is determined to be due to inconsistencies in raw material / blank material or batch material defects.

[0266] If an abnormality recurs after repair, it is determined that the initial repair was incomplete and the repaired part failed twice.

[0267] Based on process behavior, operating condition disturbance, spatiotemporal location, and component wear, multiple fluctuation sources are cross-validated, and the fluctuation sources that pass the validation are classified and labeled to obtain the fluctuation source type.

[0268] Cross-validation is performed by integrating four dimensions: process behavior, working condition disturbance, spatiotemporal location, and component wear to eliminate misjudgments. If the time and workstation of severe disturbance, process abnormality, and parameter exceeding the standard completely overlap, it is identified as a first-level core fluctuation source.

[0269] The anomalies exhibit a clear sequence: the preceding nodes experience fluctuations first, and the subsequent processes are affected in turn. The complete anomaly propagation chain has been traced.

[0270] The occurrence of similar anomalies at multiple locations independently indicates a common source, rather than a single, isolated issue.

[0271] All identified sources of fluctuation are categorized and labeled as follows: equipment, materials, environment, human factors, process flow, and component wear and tear.

[0272] Based on the type of fluctuation source and the number of disturbed processes, traceability levels are divided, traceability directions for different scenarios are determined, and a manufacturing status traceability strategy is formulated for disturbed processes and fluctuation sources in combination with a dynamic traceability adaptation mechanism.

[0273] Divide the traceability levels and clarify the traceability direction for different scenarios: Potential anomalies that only trigger warning thresholds: Focus on monitoring and tracing, track the development trend of fluctuations, temporarily avoid a complete shutdown, and focus on recording data changes.

[0274] For general anomalies, exceeding warning levels, or not yet failing to meet standards: focus on local source tracing, targeting the currently disrupted process and corresponding equipment, and investigate localized issues.

[0275] Serious anomalies reaching the non-compliance threshold: The main approach is to trace back the entire chain, starting from the current abnormal node and tracing back to all previous processes, materials, and equipment records.

[0276] Fatal flaw, exceeding the limit threshold: Initiate full batch and full domain traceability to lock down all products in the same batch, upstream and downstream supply chain materials, and the status of all production line equipment.

[0277] Invoke the dynamic traceability adaptation mechanism to match the corresponding traceability method: Single-point device anomaly: Enable fixed-point retrieval traceability, lock a single spatial grid, a single device, and the corresponding time period data.

[0278] Disruption to the process chain: Enable forward and reverse linkage retrieval to retrieve data bidirectionally along the process flow.

[0279] Cross-partition common anomalies: Batch search by functional partition, compare data of all workstations and equipment within the partition.

[0280] Batch material anomalies: Trace the batch by material batch code and link it to upstream suppliers and incoming material inspection records.

[0281] Set up a traceability data retrieval list: clearly define the content that must be extracted, including process behavior records, full-scale operating parameters, disturbance analysis reports, dynamic threshold change records, loss data, operator records, and quality inspection data.

[0282] Based on the type of fluctuation source and the characteristics of the disturbed process, a manufacturing status traceability strategy can be formulated, which may include: Scenario 1: Equipment operating condition disturbance is the core source of fluctuation, and a single process is disturbed.

[0283] Key areas for traceability: historical equipment maintenance records, sensor status, parameter setting records, and equipment inspection data.

[0284] Actions to be performed: Shut down the machine to check for equipment malfunctions, verify sensor accuracy, and recalibrate process parameters.

[0285] Traceability closed loop: Mark the device as a key monitoring object and increase the frequency of subsequent operating condition data collection and threshold verification.

[0286] Scenario 2: The process flow / human operation is the source of fluctuation, and multiple processes are disturbed in a chain reaction.

[0287] Key areas for traceability: process operation logs, personnel schedules, transfer and handover records, and process execution standards.

[0288] Actions to be taken: Review the points where violations occurred, standardize the workflow, and conduct a review of personnel operations.

[0289] Traceability closed loop: Add operation verification nodes to the abnormal process grid to strengthen process control.

[0290] Scenario 3: Component wear / raw material defects are the source of fluctuations, causing continuous anomalies throughout the entire process.

[0291] Key traceability elements: incoming material documents, raw material testing reports, full-process data of the same batch of products, and loss evolution curves.

[0292] Actions to be taken: Isolate the problematic batch of materials, conduct random inspections of work-in-process, and assess the remaining usable life of components.

[0293] Traceability closed loop: Trace the supply chain upstream and simultaneously optimize the incoming material inspection standards at the front end.

[0294] Scenario 4: Common external sources of fluctuation, such as workshop environment and power grid, cause large-scale disturbances in different areas.

[0295] Key areas of focus: zoned environmental monitoring data, power supply records, and operational status of public auxiliary equipment.

[0296] Actions to be taken: Investigate public facility malfunctions and take corrective measures such as shielding, voltage stabilization, and temperature and humidity control.

[0297] Traceability closed loop: Add zoned environmental data collection items and include them in the daily traceability monitoring scope.

[0298] Based on the same general inventive concept, this invention also protects a device manufacturing status traceability system based on IoT data acquisition, comprising: The full-domain data acquisition and sensing module is used to collect equipment operating condition data, component unique identification data, and production process behavior data during the equipment manufacturing process. Based on the component unique identification data, the manufacturing traceability boundary range is defined, and combined with the production process behavior data, the current manufacturing flow status is determined.

[0299] The retrieval mechanism adaptation module is used to set the traceability retrieval mechanism by combining the state adaptation retrieval method with the current manufacturing flow status. Based on the preset time-series grid algorithm, the traceability retrieval mechanism is modified according to the production workstation partition scenario to obtain a dynamic traceability adaptation mechanism.

[0300] The disturbance analysis modeling module is used to generate operating condition disturbance analysis data based on the degree of interference of equipment operating condition data on the steady-state parameters of the manufacturing process. It works in conjunction with the dynamic traceability adaptation mechanism and production process behavior data to build a reliable manufacturing state profile. It also combines the component wear evolution law with the operating condition disturbance analysis data to formulate dynamic quality judgment thresholds.

[0301] The linkage traceability matching module is used to perform linkage traceability matching analysis on the reliable manufacturing status profile based on the dynamic quality judgment threshold, identify the source of abnormal fluctuations and the disturbed process, and generate a manufacturing status traceability strategy.

[0302] In summary, this embodiment provides a method and system for tracing the manufacturing status of equipment based on IoT data collection. Through boundary delineation and dynamic optimization of anomalies, it defines the scope of traceability data while adapting to complex production scenarios, ensuring continuous and uninterrupted traceability throughout the entire process and significantly improving the effectiveness of traceability data. By subdividing process status into various types and benchmarking against standard processes, it can promptly identify issues such as skipped steps, disordered sequences, and violations, achieving refined control of the production process and standardizing on-site operations. The retrieval rules automatically adjust according to the manufacturing flow status, workstation partitioning, and temporal scenarios. Different processes and regions employ differentiated retrieval logic, data granularity, and display formats, resulting in faster retrieval response and more accurate filtering, adapting to complex manufacturing sites with multiple workstations and processes. It fully considers the natural wear and tear evolution of parts and the cumulative impact of operational disturbances. Thresholds are automatically updated according to the wear stage, disturbance intensity, and production events, avoiding the misclassification of normal parts as defective products while tightening standards during strong disturbances and critical wear stages to prevent defective products from being released, significantly improving the accuracy of quality assessment.

[0303] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0304] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for tracing the manufacturing status of equipment based on IoT data acquisition, characterized in that, include: Collect equipment operating condition data, component unique identification data and production process behavior data during the equipment manufacturing process; define the manufacturing traceability boundary range based on the component unique identification data; and determine the current manufacturing flow status by combining the production process behavior data. A state-adaptive retrieval method is adopted to set a traceability retrieval mechanism in combination with the current manufacturing flow status. Based on a preset temporal grid algorithm, the traceability retrieval mechanism is modified according to the production workstation partition scenario to obtain a dynamic traceability adaptation mechanism. The steps to obtain the dynamic traceability adaptation mechanism include: According to a preset unified time step, the entire production cycle of the parts is divided into a time-series grid. The physical space of the workshop is divided into corresponding spatial grids using a preset minimum spatial unit. The time-series grids and the spatial grids are combined to form spatiotemporal grid nodes. Combined with the production process, a preset flow route is generated. The workshop is divided into functional zones according to its operational functions, which are then bound to the spatial grid, and production characteristics are recorded as a retrieval basis. The traceability retrieval mechanism is mapped onto the spatiotemporal grid nodes. Based on the actual process behavior records of the parts, the spatiotemporal grid nodes that have passed through are matched to obtain the actual flow route. The trajectory deviation between the actual flow route and the preset flow route is calculated from three aspects: timing deviation, spatial deviation, and state deviation, and the correction steps are determined. Based on the search criteria and the correction steps, the search initiation threshold and triggering rules are corrected, the search data filtering range is corrected based on the spatial grid, and the data retrieval granularity rules are corrected based on the temporal grid. Based on the partition scenario and grid status, the logic for switching between cross-scenario searches was revised, and the display format of search results was revised based on the requirements of partitioned tasks; Based on the degree of interference of the equipment operating condition data to the steady-state parameters of the manufacturing process, operating condition disturbance analysis data is generated. The dynamic traceability adaptation mechanism and the production process behavior data are used to construct a reliable manufacturing state profile. Dynamic quality judgment thresholds are formulated by combining the component wear evolution law with the operating condition disturbance analysis data. Based on the dynamic quality judgment threshold, a linked traceability matching analysis is performed on the reliable manufacturing status profile to identify abnormal fluctuation sources and disturbed processes, thereby generating a manufacturing status traceability strategy.

2. The method for tracing the manufacturing status of equipment based on IoT data acquisition according to claim 1, characterized in that: The steps for defining the manufacturing traceability boundary include: The fields are disassembled according to the preset coding rules, and the basic coding information in the unique identification data of the parts is extracted. The traceability query needs are determined based on the currently queried parts. Based on the traceability query requirements, hierarchical traceability boundaries are defined, the spatiotemporal information of the current component identifier is associated, spatial boundaries are delineated, the production timeline corresponding to the identifier is matched, and time boundaries are defined. Matching identifiers to bind processes, defining process flow boundaries, associating identifier-bound equipment operating condition data to determine data boundaries, and dynamically adjusting the hierarchical traceability boundaries, spatial boundaries, temporal boundaries, process flow boundaries, and data boundaries based on different abnormal operating conditions to obtain the manufacturing traceability boundary range.

3. The method for tracing the manufacturing status of equipment based on IoT data acquisition according to claim 2, characterized in that: The steps to determine the current manufacturing flow status include: Based on the coding information, an ordered process node table is obtained by associating the corresponding product process route and bill of materials, and a standard process node table is formed by combining the manufacturing traceability boundary range. Using the current component identifier as the primary key, extract the corresponding component's behavior events from the process behavior data and arrange them in ascending order by timestamp to form an event chain; The event chain is aligned and matched with the standard process node table one by one, and the status is classified to obtain the current process node of the component. The current manufacturing flow status is obtained by identifying the flow direction and flow status of process nodes from forward flow, rework, temporary storage and waiting, and cross-line transfer.

4. The method for tracing the manufacturing status of equipment based on IoT data acquisition according to claim 1, characterized in that: The steps for setting up the traceability retrieval mechanism include: Define search requirements corresponding to different manufacturing flow states, use the current component identifier as the search primary key, and build a search index system in combination with the manufacturing traceability boundary range; For each type of manufacturing flow status, set corresponding search start thresholds and trigger rules, and set multiple search modes according to the differences in flow status; Once the retrieval index system is activated, data is filtered according to the three-tiered order of manufacturing traceability boundary range, process stage, and data type. Based on the current manufacturing flow status, match the types of data that need to be retrieved, set multiple display levels according to the query depth requirements corresponding to the flow status, and reset the search request and re-execute the search when the manufacturing flow status changes.

5. The method for tracing the manufacturing status of equipment based on IoT data acquisition according to claim 1, characterized in that: The steps for generating the operating condition disturbance analysis data include: Establish a manufacturing steady-state benchmark parameter system corresponding to different processes, and filter the effective operating condition data within the manufacturing traceability boundary from the equipment operating condition data; The scenario in which the actual operating parameters of the equipment deviate from the manufacturing steady-state reference parameter system is defined as a working condition disturbance, and different disturbance levels and disturbance forms are divided to obtain a disturbance judgment classification rule. For each data point in the effective operating condition data, the interference index is calculated by comparing them one by one based on the manufacturing steady-state benchmark parameter system, and the weights are set according to the process characteristics to calculate the interference score. The disturbance scores of all valid operating condition data are weighted and summarized according to the weight ratio to obtain the current disturbance coefficient. The location, triggering cause and disturbance type of the disturbance are located to obtain the disturbance analysis data of the operating condition.

6. The method for tracing the manufacturing status of equipment based on IoT data acquisition according to claim 1, characterized in that: The steps for constructing the trusted manufacturing state profile include: Using the unified timestamp of the time-series grid as a benchmark, the occurrence time of the process behavior, the start and end time of the working condition disturbance, and the data acquisition time are calibrated and aligned, and then attached to the current component identifier to establish the correspondence between identifier, process behavior, working condition disturbance, and grid partition. Based on the needs of production control and traceability verification, determine the dimensions of profile creation and the corresponding data content to be filled in, and build a profile framework; Based on the dynamic traceability and adaptation mechanism, matching data is retrieved according to the profile dimensions, and the details of the work flow status are extracted according to the production process behavior data. The working condition disturbance analysis data is then superimposed to supplement the stable operating status of the equipment. By combining the spatiotemporal grid nodes with the retrieval criteria, the manufacturing trajectory status is improved, and behavioral perturbation and spatiotemporal status verification are performed. For the aforementioned portrait characterization dimensions, quantitative evaluation indicators are set, and the manufacturing state is characterized by a combination of qualitative description and quantitative values. Based on the disturbance level and process abnormality, the overall state level is divided. The content is filled in sequentially according to the portrait framework, and bound to the dynamic traceability and adaptation mechanism to dynamically update the portrait and obtain the trusted manufacturing status portrait.

7. The method for tracing the manufacturing status of equipment based on IoT data acquisition according to claim 1, characterized in that: The steps for determining the dynamic quality judgment threshold include: Collect component loss data, break down different evolution stages according to processing sequence and cumulative loss level, determine the loss characteristics and inherent loss baseline of each stage, and the weight of the impact of production process on component loss. Based on the operating condition disturbance analysis data, the impact of disturbances on losses is classified and matched. For each type of loss and its evolution stage, the disturbance impact coefficient is calculated, and a disturbance loss correlation library is constructed. Based on the technical requirements and inspection specifications of the components, the judgment indicators are determined, and the basic quality judgment threshold is set according to the inherent loss baseline and drawing standards to distinguish the judgment criteria at different levels. Identify the current evolution stage, correct the basic quality judgment threshold based on the loss characteristics and influence weights, and extract the disturbance influence coefficient corresponding to the current disturbance level from the disturbance loss association library for secondary correction to generate a temporary dynamic threshold; The temporary dynamic threshold is divided into threshold effective periods, and switching trigger conditions are set based on evolution stage jumps, disturbance level changes, process switching, and processing batch replacements.

8. The method for tracing the manufacturing status of equipment based on IoT data acquisition according to claim 6, characterized in that: The steps for generating the manufacturing status traceability strategy include: Extract the measured data of each process period in the manufacturing status profile, compare each data point with the current dynamic quality judgment threshold, classify and mark the quality anomaly types, and sort all the abnormal nodes in chronological order to form an abnormal time interval. Align the abnormal time interval with the production process behavior data along the time axis to locate the disturbed process corresponding to the quality anomaly type. For the disturbed process, the abnormal fluctuation source is located by combining the working condition disturbance analysis data, and the spatial grid nodes and work station partition information are linked to investigate spatial flow type fluctuation sources, and the body type fluctuation source is traced by combining the evolution stage of the component. Based on process behavior, operating condition disturbance, spatiotemporal location and component wear, multiple fluctuation sources are cross-validated, and the fluctuation sources that pass the validation are classified and labeled to obtain the fluctuation source type. Based on the type of fluctuation source and the number of disturbed processes, traceability levels are divided, traceability directions for different scenarios are determined, and the manufacturing status traceability strategy is formulated for disturbed processes and fluctuation sources in conjunction with the dynamic traceability adaptation mechanism.

9. A device manufacturing status traceability system based on IoT data acquisition, applied to a device manufacturing status traceability method based on IoT data acquisition as described in any one of claims 1 to 8, characterized in that, The traceability system includes: The full-domain acquisition and perception module is used to collect equipment operating condition data, component unique identification data and production process behavior data during the equipment manufacturing process. Based on the component unique identification data, the manufacturing traceability boundary range is defined, and combined with the production process behavior data, the current manufacturing flow status is determined. The retrieval mechanism adaptation module is used to set a traceability retrieval mechanism by combining the current manufacturing flow status with the state adaptation retrieval method, and to modify the traceability retrieval mechanism according to the production station partitioning scenario based on the preset temporal grid algorithm to obtain a dynamic traceability adaptation mechanism. The disturbance analysis modeling module is used to generate operating condition disturbance analysis data based on the degree of interference of the equipment operating condition data to the steady-state parameters of the manufacturing process, and to construct a reliable manufacturing state profile in conjunction with the dynamic traceability adaptation mechanism and the production process behavior data. It also combines the component loss evolution law with the operating condition disturbance analysis data to formulate a dynamic quality judgment threshold. The linkage traceability matching module is used to perform linkage traceability matching analysis on the reliable manufacturing status profile based on the dynamic quality judgment threshold, identify abnormal fluctuation sources and disturbed processes, and generate manufacturing status traceability strategies.

Images