Real-time production management and control method and system based on incremental rearrangement

Abstract

The present application relates to the technical field of production system automation control, and provides a real-time production management and control method and system based on incremental rearrangement.The method comprises the following steps: listening to target events in a manufacturing system and classifying and identifying the target events, determining whether to trigger a scheduling recalculation request; identifying a process set having a process dependency relationship or a resource conflict relationship with the target events, and constructing a local scheduling domain; generating a local recalculation data set based on the local scheduling domain, and setting a frozen identifier and / or a write control field for processes outside the local scheduling domain; performing effective processing time window allocation, conflict detection and incremental sorting rearrangement within the local scheduling domain; and limiting the updated production scheduling result to be written back into the local scheduling domain.The present application can avoid non-associated processes from being updated or rearranged together, and improve production scheduling response speed and plan stability.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing production scheduling, and in particular to a real-time production control method and system based on incremental rearrangement. Background Technology

[0002] In manufacturing systems of discrete manufacturing industries involving multiple parts, production plans are often affected by high-frequency disturbances such as order changes, equipment malfunctions, process delays, quality problems, human intervention, and approval adjustments, and need to be scheduled through a production scheduling system.

[0003] However, existing production scheduling systems mostly rely on global task queues for scheduling recalculation. When a local target event occurs, the entire production plan often needs to be recalculated. Existing production scheduling systems suffer from high computational complexity, large response latency, poor plan stability, disconnect between management events and scheduling, and insufficient adaptation to non-standard processes. Summary of the Invention

[0004] Based on this, the purpose of the present invention is to provide a real-time production control method and system based on incremental rearrangement, which can respond in real time after an event occurs and accurately limit the affected local scheduling scope, avoiding unrelated processes from being updated or rearranged, thereby achieving localized control of the production process.

[0005] A real-time production control method based on incremental rearrangement includes the following steps:

[0006] S100: Monitor target events in the manufacturing system and classify and identify the target events; based on the category of the target events, determine whether a scheduling recalculation request is triggered. If a scheduling recalculation request is triggered, proceed to step S200; otherwise, continue monitoring. S200. Identify the set of processes that have process dependencies or resource conflicts with the target event, and construct a local scheduling domain; set a freeze flag and / or write control fields to processes outside the local scheduling domain, so that processes outside the local scheduling domain do not participate in production scheduling calculation in the current scheduling cycle, and their production scheduling records are not updated or rearranged by the production scheduling calculation results of the current scheduling cycle. S300. Based on the local scheduling domain, obtain the data to be recalculated, and extract, generate or filter the local recalculation dataset from the data to be recalculated; the set of processes corresponding to the local recalculation dataset does not exceed the local scheduling domain. S400. Based on the local recalculation dataset, within the local scheduling domain, the effective processing time window is proportionally allocated, and the theoretical start time and theoretical end time of each process are calculated. S500. Based on the local recalculation dataset, detect time conflicts and resource conflicts within the local scheduling domain, and obtain a subset of conflicting procedures. S600. Based on the multi-rule priority model, perform incremental sorting and rearrangement on the conflicting process subset, and update the incremental scheduling result of the local scheduling domain. S700: Write back the updated scheduling results, and limit the write-back scope to the local scheduling domain based on the local domain identifier and / or write control field, so that the scheduling records of processes outside the local scheduling domain are not written, overwritten, updated or rearranged.

[0007] The real-time production control method based on incremental reordering described in this invention constructs a local scheduling domain by responding to target events, limiting the scope of scheduling calculations, conflict resolution, incremental reordering, and writing back of scheduling results for sorting optimization. This enables subsequent conflict detection and incremental reordering to accurately target the processes and resources associated with the target events, improving the response speed and plan stability of production scheduling.

[0008] Further, in step S100, the target event includes at least one of the following: order change event, equipment status change event, process status change event, quality anomaly event, manual intervention event, and approval status change event.

[0009] Further, after determining the triggering of the scheduling recalculation request in step S100, the triggered scheduling recalculation request is detected; if the scheduling recalculation request points to the global production scheduling recalculation of all orders or all processes, the scheduling recalculation request is intercepted, blocked, or blocked at the entry point; if the scheduling recalculation request does not point to the global production scheduling recalculation of all orders or all processes, then proceed to step S200. Further, in step S200, based on the pre-established and maintained process dependency graph and resource occupation graph, the affected edges or nodes are updated in conjunction with the target event, and within the range of satisfying at least one of the graph traversal depth threshold and resource conflict hop count threshold, the set of processes that have process dependency relationship or resource conflict relationship with the target event is identified.

[0010] Furthermore, in non-standard discrete manufacturing scenarios where a Bill of Materials (BOM) is missing or the process route changes dynamically, a process sequence and its sequential constraints are generated based on the order-process mapping relationship, the process rule base, and / or the real-time collected work-in-process status data. A process dependency graph is constructed or updated, and the local scheduling domain is constructed based on the updated process dependency graph.

[0011] Furthermore, step S200 also includes limiting the maximum number of affected processes and filtering out non-overlapping processes based on the overlap between the target event time window and the process planning time window, thereby further narrowing the local scheduling domain.

[0012] Further, in step S300, the data to be recalculated includes at least one of order data, process data, resource usage data, calendar data, and work-in-process status data; the local recalculation dataset is generated through an event-process mapping index, a process-resource inverted index, a change log incremental cursor, and / or a process dependency closure query. Freezing the scheduling results of processes not included in the local recalculation dataset includes setting a freeze flag for the processes outside the local scheduling domain and / or writing it into a control field; Furthermore, it also includes setting local domain identifiers and / or writing control fields for scheduling recalculation requests and write-back requests, and performing consistency checks on the scheduling recalculation requests and write-back requests in subsequent steps; during the scheduling recalculation process triggered by the target event, a local scheduling recalculation channel is established, and global scheduling recalculation requests pointing to all orders or all processes are intercepted, blocked, or blocked at the entry point. If it is detected that the scheduling recalculation request or write-back request does not carry a local domain identifier, the range of the local recalculation dataset exceeds the local scheduling domain, the write-back target includes processes outside the local scheduling domain, or the scheduling call chain has a processing flow of first performing global scheduling recalculation and then filtering local results, it is determined that the consistency check has failed. When the consistency check fails, the scheduling recalculation request or write-back request is rejected, the global production scheduling recalculation entry is blocked, or the scheduling recalculation request or write-back request is downgraded to only update the business status without triggering production scheduling calculation.

[0013] Further, step S400 includes obtaining an effective processing time window within the local scheduling domain after deducting non-production periods based on the local recalculation dataset, and allocating the effective processing time window proportionally based on the theoretical working time weight of each process, and calculating the theoretical start time and theoretical end time of each process.

[0014] Further, step S400 also includes collecting the theoretical working time and actual processing time of the process, and generating working time deviation analysis results, correction reference information and / or candidate parameter adjustment suggestions based on the deviation between the theoretical working time and the actual processing time. The theoretical working time weight and the theoretical working time of the process are confirmed and adjusted manually. The theoretical working time weight generates candidate correction coefficients through an error feedback correction model. The error feedback correction model statistically analyzes the deviation between the theoretical working time and the actual processing time based on the process dimension and / or resource dimension, and generates the candidate correction coefficients according to a preset time period and / or according to event triggering. After manual confirmation, the theoretical working time weight and / or the theoretical working time parameter of the process are updated based on the candidate correction coefficients.

[0015] Furthermore, the multi-rule priority model includes one or more priority rules; the priority rules include at least one or more of the following: order delivery date priority, critical path weight of the process, resource bottleneck weight, work-in-process waiting time weight, and manual intervention priority; the incremental sorting and rearrangement includes generating a priority sequence for processes with time conflicts and / or resource conflicts, and based on the priority sequence, performing local adjustments on the conflicting process subset using an incremental solution method to resolve the conflicts; the local adjustments include at least one or more of the following: local exchange, local insertion, and local movement, and the incremental sorting and rearrangement is limited to the conflicting process subset and its adjacent affected processes.

[0016] Based on the same inventive purpose, embodiments of this application also provide a real-time production control system based on incremental reordering, including an event monitoring and classification module, a local scheduling domain construction module, a local recalculation dataset control module, a time window ratio calculation module, a conflict detection and incremental sorting module, a pseudo-incremental protection execution control module, and a write-back module.

[0017] The event monitoring and classification module is used to monitor target events in the manufacturing system, classify and identify target events, and determine whether to trigger a scheduling recalculation request based on the category of the target event.

[0018] The local scheduling domain construction module is used to identify a set of processes that have process dependencies or resource conflicts with the target event, and to construct a local scheduling domain.

[0019] The local recalculation dataset control module is used to extract, generate, or filter local recalculation datasets from the data to be recalculated based on the local scheduling domain. The local recalculation dataset control module is also used to set a freeze flag and / or write control fields to processes outside the local scheduling domain, so that processes outside the local scheduling domain do not participate in production scheduling calculations in the current scheduling cycle, and their production scheduling records are not updated or rearranged by the production scheduling calculation results of the current scheduling cycle.

[0020] The time window ratio calculation module is used to allocate effective processing time windows proportionally within the local scheduling domain based on the local recalculation dataset, and to calculate the theoretical start time and theoretical end time of each process.

[0021] The conflict detection and incremental sorting module is used to detect time conflicts and resource conflicts in the local scheduling domain based on the local recalculation dataset, obtain a subset of conflicting processes, and perform incremental sorting and rearrangement of the subset of conflicting processes based on a multi-rule priority model to update the incremental scheduling results of the local scheduling domain.

[0022] The pseudo-incremental protection execution control module is used to set local domain identifiers and / or write control fields for scheduling recalculation requests and write-back requests, and to perform consistency checks on the scheduling recalculation requests and write-back requests. If it is detected that the scheduling recalculation request or write-back request does not carry a local domain identifier, the range of the local recalculation dataset exceeds the local scheduling domain, the write-back target includes a process outside the local scheduling domain, or the scheduling call chain has a process of first performing global production scheduling recalculation and then filtering local results, it is considered that the consistency check has failed. When the consistency check fails, the pseudo-incremental protection execution control module refuses to execute the scheduling recalculation request or write-back request, blocks the global production scheduling recalculation entry, or downgrades the scheduling recalculation request or write-back request to only update the business status without triggering production scheduling calculation.

[0023] The write-back module is used to write back the updated production scheduling results and control the write-back range of the production scheduling results based on the local domain identifier and / or write control field, so that the production scheduling records of processes outside the local scheduling domain are not written, overwritten, updated or rearranged.

[0024] To better understand and implement this invention, the following detailed description is provided in conjunction with the accompanying drawings. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a schematic diagram of a real-time production control method based on incremental rearrangement according to an embodiment of the present invention. Detailed Implementation

[0027] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0028] In this invention, unless otherwise explicitly defined, the terms "module," "unit," "field," "identifier," "request," "channel," "dataset," "local scheduling domain," "write-back," "freeze," and "reordering," etc., should be interpreted broadly. For example, "module" or "unit" can be a functional component, business logic unit, data processing unit, or a combination thereof that implements the corresponding function; "field" or "identifier" can be a database field, business status field, control parameter, tag, or other data marker that can be used to define the scope of processing; "write-back" can refer to the addition, updating, or overwriting control of production scheduling records, process status, resource occupancy records, or scheduling result records. Those skilled in the art can understand the meaning of the above terms based on the specific technical solutions disclosed in this invention.

[0029] In non-standard discrete manufacturing scenarios, the production system continuously operates and handles a variety of high-frequency disturbance events, including new order insertion, equipment malfunctions, process delays, quality rework, manual order insertion, and approval adjustments.

[0030] Traditional methods typically involve a global recalculation of production schedules for all orders or processes upon each event trigger, resulting in high computational latency. Furthermore, a single localized event can easily trigger a chain reaction of adjustments to the planned times of numerous unrelated processes, causing schedule instability. While some systems handle disturbance events under the guise of local scheduling, their actual processing flow involves first performing a global recalculation and then filtering local results from the global schedule. This approach still incurs significant computational overhead and may lead to the updating or overwriting of schedules for unrelated processes.

[0031] Example Please see Figure 1 The present invention provides a real-time production control method based on incremental rearrangement, which specifically includes the following steps: S100: Monitor target events in the manufacturing system and classify and identify the target events; based on the category of the target events, determine whether a scheduling recalculation request is triggered. If a scheduling recalculation request is triggered, proceed to step S200; otherwise, continue monitoring.

[0032] S200. Identify the set of processes that have process dependencies or resource conflicts with the target event, and construct a local scheduling domain; set a freeze flag and / or write control fields to processes outside the local scheduling domain, so that processes outside the local scheduling domain do not participate in production scheduling calculation in the current scheduling cycle, and their production scheduling records are not updated or rearranged by the production scheduling calculation results of the current scheduling cycle.

[0033] S300. Based on the local scheduling domain, obtain the data to be recalculated, and extract or filter the data to be recalculated to obtain a local recalculation dataset; the set of processes corresponding to the local recalculation dataset does not exceed the local scheduling domain.

[0034] S400. Based on the local recalculation dataset, within the local scheduling domain, the effective processing time window is proportionally allocated, and the theoretical start time and theoretical end time of each process are calculated.

[0035] S500. Based on the local recalculation dataset, detect time conflicts and resource conflicts within the local scheduling domain, and obtain a subset of conflicting procedures.

[0036] S600. Based on the multi-rule priority model, perform incremental sorting and rearrangement on the subset of conflicting processes, and update the incremental scheduling result of the local scheduling domain.

[0037] S700: Write back the updated scheduling results. Based on the local domain identifier and / or write control field, limit the write-back scope to the local scheduling domain, so that the scheduling records of processes outside the local scheduling domain are not written, overwritten, updated, or rearranged.

[0038] The real-time production control method based on incremental rescheduling provided in this application first constructs a local scheduling domain based on a target event, and generates a local recalculation dataset only based on the local scheduling domain, ensuring that the set of processes corresponding to the local recalculation dataset does not exceed the local scheduling domain. Within the current scheduling cycle, the system sets a freeze flag and / or writes a control field for processes outside the local scheduling domain, preventing these processes from participating in production scheduling calculations and ensuring that their scheduling records are not updated or rescheduled by the production scheduling calculation results of the current scheduling cycle. Simultaneously, the system limits the scope of writing back the scheduling results to the local scheduling domain through the local domain flag and / or write control field, preventing the scheduling records of processes outside the local scheduling domain from being written, overwritten, updated, or rescheduled, thereby avoiding the cascading adjustments of unrelated processes and improving production scheduling response speed and plan stability.

[0039] In step S100, the target events include, but are not limited to, order change events, equipment status change events, process status change events, quality anomaly events, manual intervention events, or approval status change events; order change events include, but are not limited to, new order entry events, order delivery date change events, order processing requirement change events, or order priority change events.

[0040] Furthermore, step S100 also includes detecting the triggered scheduling recalculation request. If the scheduling recalculation request points to the global production recalculation of all orders or all processes, the scheduling recalculation request is intercepted, blocked, or blocked at the entry point to prevent the triggering of the global production recalculation of all processes or all orders.

[0041] Further, in step S200, identifying the set of processes that have process dependencies or resource conflicts with the target event specifically includes: the system, based on a pre-established and maintained process dependency graph and resource occupancy graph, updates affected edges or nodes in conjunction with the target event, and identifies the set of affected processes that have process dependencies or resource conflicts with the target event within a limited range that satisfies at least one of the graph traversal depth threshold and the resource conflict hop count threshold. By constructing a local scheduling domain by limiting the graph traversal level, the maximum resource conflict hop count, or the maximum number of affected processes, the complexity of local scheduling recalculation can be controlled.

[0042] Furthermore, step S200 also includes limiting the maximum number of affected processes and filtering out non-overlapping processes based on the overlap between the target event time window and the process planning time window, so as to further narrow down the local scheduling domain.

[0043] The time window for a target event is the time interval between the occurrence of the target event and the duration of its impact. The time window for a process plan is the time interval between the planned start time and the planned end time of the process. Furthermore, in non-standard discrete manufacturing scenarios where there is a missing Bill of Materials (BOM) or dynamic changes in the process route, the process sequence and its sequential constraints are generated based on the order-process mapping relationship, the process rule base, and / or the real-time collected work-in-process status data. The process dependency graph is constructed or updated, and the local scheduling domain is constructed based on the updated process dependency graph.

[0044] In step S300, the data to be recalculated includes at least one of order data, process data, resource usage data, calendar data, and work-in-process status data.

[0045] The local recalculation dataset is generated by using event-process mapping indexes, process-resource inverted indexes, change log incremental cursors, and / or process dependency closure queries.

[0046] Furthermore, in step S300, the system also sets local domain identifiers and / or writes them into control fields for scheduling recalculation requests and write-back requests, and performs consistency checks on scheduling recalculation requests and write-back requests in subsequent steps. If it is detected that the request does not carry a local domain identifier, the range of the local recalculation dataset exceeds the local scheduling domain, or the write-back target includes a process outside the local scheduling domain, or the scheduling call chain has a process of first performing global production recalculation and then filtering local results, it is considered that the consistency check has failed. When the consistency check fails, the system refuses to execute the scheduling recalculation request or write-back request, blocks the global production recalculation entry, or downgrades the scheduling recalculation request or write-back request to only update the business status, so as to prevent local incremental scheduling from degenerating into a pseudo-incremental scheduling path.

[0047] Furthermore, step S400 specifically includes: based on the local recalculation dataset, within the local scheduling domain, after deducting non-production periods within the order delivery time window, obtaining an effective processing time window, and proportionally allocating the effective processing time window based on the theoretical working time weight of each process, so as to calculate the theoretical start time and theoretical end time of each process.

[0048] Furthermore, the non-production period includes periods when resources are unavailable and / or production line downtime periods; wherein, the period when resources are unavailable refers to the period during which resources are unprocessable within the order delivery time window; and the period during which production line downtime refers to the period during which the production line stops production within the order delivery time window.

[0049] The reasons for non-performance of processing or non-production include at least maintenance, mold change, upkeep, malfunction, shift scheduling, holidays, or planned line stoppages.

[0050] Furthermore, step S400 also includes collecting the theoretical working time and actual processing time of the process, and generating working time deviation analysis results, correction reference information and / or candidate parameter adjustment suggestions based on the deviation between the theoretical working time and the actual processing time, and manually confirming and adjusting the theoretical working time weight and the theoretical working time of the process.

[0051] In a preferred embodiment provided in this application, the method for feedback correction of the deviation between theoretical working hours and actual processing time is as follows: The system determines the order delivery time window based on the order delivery date, and then deducts periods of resource unavailability and / or production line downtime within the order delivery time window to obtain the effective processing time window. The system proportionally allocates the effective processing time window according to the theoretical time weight of each process to determine the theoretical start time and theoretical end time of each process.

[0052] During the execution of a process, the system collects data such as the actual start time, actual end time, actual processing duration, resources used, operators, processing specifications, and processing status. It then compares the actual processing duration with the theoretical processing time to generate deviation data between the theoretical and actual processing times.

[0053] The system performs statistical analysis on the deviation data based on the process dimension and / or resource dimension, generating time deviation analysis results, correction reference information, and / or candidate parameter adjustment suggestions. The time deviation analysis results may include the average deviation, maximum deviation, deviation ratio, or deviation trend of a certain process within a preset time period; the correction reference information may include the theoretical time weights to be adjusted, candidate correction coefficients, or abnormal deviation prompts.

[0054] In this embodiment, the proposed candidate parameter adjustments do not automatically overwrite the existing theoretical time parameters of the process. Instead, they are provided for manual confirmation and adjustment of the theoretical time weights and / or the theoretical time parameters of the process. The manually confirmed parameters can be written into the system and used as reference input for subsequent local scheduling domain production.

[0055] Through the above methods, the system can continuously adjust the production scheduling parameters based on actual production data, thereby improving the consistency between subsequent local incremental production scheduling results and actual production conditions.

[0056] The theoretical working time weights are used to generate candidate correction coefficients through an error feedback correction model. This model statistically analyzes the deviation between theoretical and actual processing times based on process and / or resource dimensions, and generates the correction coefficients according to a preset time period and / or event triggering. After manual confirmation, the theoretical working time weights and / or process theoretical working time parameters are updated based on the candidate correction coefficients, and the updated parameters are used as subsequent reference inputs to improve the consistency between production scheduling results and actual production conditions.

[0057] In steps S500 and S600, the system detects time and resource conflicts within the local scheduling domain and identifies the processes with detected time and / or resource conflicts as a subset of conflicting processes. The system generates a priority sequence for the processes in the subset of conflicting processes based on a multi-rule priority model, and performs local adjustments to the subset of conflicting processes using an incremental solution method based on the priority sequence to resolve the conflicts. The multi-rule priority model includes one or more of the following: order delivery date priority, process critical path weight, resource bottleneck weight, work-in-process waiting time weight, and manual intervention priority. The local adjustments include one or more of the following: local exchange, local insertion, and local movement, and the incremental sorting rearrangement is limited to the subset of conflicting processes and their adjacent affected processes.

[0058] In step S700, the system only writes the updated production scheduling results back to the corresponding production scheduling records within the local scheduling domain, and controls the write-back range of the production scheduling results based on the local domain identifier and / or write control field, so that the production scheduling records of processes outside the local scheduling domain are not written, overwritten, updated or rearranged.

[0059] Example 1: Local Incremental Reordering Triggered by New Order Insertion Event When the production system receives a new order entry event, the system identifies the event as an order change event and triggers a scheduling recalculation request based on the order change event.

[0060] The system uses the processes and resources involved in the new order as seeds, and expands and identifies the set of affected processes that have process dependencies or resource conflicts with them based on the process dependency graph and resource occupation graph. Specifically, the system determines the preceding processes, succeeding processes, and other processes that share resources with the process nodes involved in the new order, and constructs a local scheduling domain under the constraints of graph traversal depth threshold, resource conflict hop count threshold, and maximum number of affected processes.

[0061] The system extracts at least one of the following from the local scheduling domain: order data, process data, resource usage data, calendar data, and work-in-process status data, to form a local recalculation dataset. For processes outside the local scheduling domain, the system sets a freeze flag and / or writes it into a control field, so that it does not participate in production scheduling calculations in the current scheduling cycle, and its production scheduling records are not updated or rescheduled.

[0062] Within the effective processing time window after deducting non-production periods, the system calculates the theoretical start and end times of each process and performs incremental sorting and rearrangement for processes with time and / or resource conflicts based on rules such as order delivery priority, resource bottleneck weight, and process critical path weight.

[0063] When writing back the production scheduling results, the system only writes back the corresponding production scheduling records within the local scheduling domain. It does not write, overwrite, update, or rearrange the production scheduling records of processes outside the local scheduling domain, thereby achieving local updates in the scenario of inserting new orders and maintaining the stability of the global plan.

[0064] Example 2: Local scheduling domain update triggered by device state change event When the production system detects equipment failure, equipment maintenance, equipment downtime, equipment recovery, or changes in equipment processing capacity, the system identifies the event as an equipment status change event and triggers a scheduling recalculation request based on the equipment status change event.

[0065] Starting with the device node whose state has changed, the system identifies the process node that is currently occupying the device, plans to occupy the device, or has a resource conflict relationship with the device based on the resource occupation graph, and further identifies the related processes that have a sequential dependency relationship with the above processes based on the process dependency graph.

[0066] When constructing a local scheduling domain, the system limits the scope of expansion along resource conflict relationships based on a resource conflict hop count threshold, limits the levels of expansion along process dependencies based on a graph traversal depth threshold, and limits the scale of processes included in the local scheduling domain based on the maximum number of affected processes. Furthermore, the system filters out processes that do not overlap with the equipment state change event in time, based on the overlap between the duration of the event's impact and the process's planned time window.

[0067] When equipment failure or downtime renders the originally planned processing period unavailable, the system will include the affected processes in the local scheduling domain and recalculate the theoretical start and end times based on the local recalculation dataset. When the equipment becomes available again, the system can perform incremental sorting and rearrangement only for processes that have overlapping time windows or resource usage relationships with the equipment recovery event.

[0068] Through the above processing, equipment status change events only affect local processes that have resource occupation, resource conflict, or process dependence relationships with the equipment, without triggering a global recalculation of production scheduling for all orders or all processes.

[0069] Example 3: Construction of Process Dependency Graph in Non-Standard Discrete Manufacturing Scenarios In non-standard discrete manufacturing scenarios where a Bill of Materials (BOM) is missing or the process route changes dynamically, the system generates process sequences and their sequential constraints based on the order-process mapping relationship, the process rule library, and / or real-time collected work-in-process status data, and constructs or updates the process dependency graph accordingly.

[0070] Specifically, when an order does not have a fixed BOM or standard process route, the system generates a candidate process sequence by calling the process rule base based on the product type, processing specifications, processing method, delivery requirements, and current work-in-process status in the order. The candidate process sequence includes one or more processing steps and their sequential constraints. The system further combines real-time collected work-in-process status data to identify the status of completed, in-process, awaiting, or paused processes, and updates the process nodes and dependency edges in the process dependency graph accordingly.

[0071] When the process route changes due to manual adjustments, equipment capacity changes, quality anomaly handling, or approval results, the system updates the process sequence and its sequential constraints based on the changed order-process mapping relationship and process rule base, and constructs or updates the process dependency graph. Subsequently, the system identifies the set of affected processes based on the updated process dependency graph and resource occupancy graph, and constructs a local scheduling domain.

[0072] In this way, even in production scenarios where a fixed BOM is missing or the process route changes dynamically, the system can still form a process dependency basis for the construction of local scheduling domains, thereby supporting local incremental rearrangement.

[0073] Example 4: Prioritized Coverage of Human Intervention Events When a planner, production manager, or other authorized operator issues an instruction to prioritize processing, insert an order, lock the processing sequence, fix the processing time, or specify resources for a target process, the system recognizes the instruction as a manual intervention event and triggers an incremental production schedule update in the corresponding local scheduling domain based on the manual intervention event.

[0074] When the target event is a manually intervened event, the system assigns a higher priority weight to the manually designated process than the automatic scheduling rules and / or sets a manual lock flag. The manual lock flag indicates that the processing order, processing time, or resources occupied by the process should be preferentially retained during the incremental sorting and rearrangement process; the manual intervention priority weight is used to ensure that the manually designated process obtains a higher sorting weight than the automatic sorting rules in the multi-rule priority model.

[0075] During the incremental solution process of incremental sorting and rearrangement, the system identifies manually specified processes and processes that conflict with it in terms of time or resources as a subset of conflicting processes. It then generates a priority sequence based on rules such as the priority weight of manual intervention, order delivery priority, resource bottleneck weight, and work-in-process waiting time weight. For processes with manually locked flags, the system prioritizes their processing position, processing time, or occupied resources during local adjustments. For processes requiring priority insertion, the system searches for insertion positions that meet the constraints within the local scheduling domain.

[0076] Through the above processing, manual intervention events can be prioritized and responded to within the local scheduling domain, while avoiding the rescheduling of all orders or all processes due to manual order insertion or manual locking.

[0077] Example 5: Handling the rework, redo, and release path for quality anomaly events When the production system detects a quality anomaly event, the system identifies the event as a quality anomaly event. These quality anomalies include, but are not limited to, abnormal processing dimensions, abnormal surface quality, processing errors exceeding thresholds, abnormal customer acceptance, quality inspection failures, or abnormal feedback from the production process.

[0078] The system determines the rework, redo, or release path based on the quality anomaly handling results. When the quality anomaly handling result is rework, the system includes the rework process, the resources required for rework, and processes with dependencies or resource conflicts with the rework process in the local scheduling domain, and performs local recalculation and incremental sorting rearrangement based on the local recalculation dataset. When the quality anomaly handling result is redo, the system includes the redo process and its subsequent affected processes in the local scheduling domain, and recalculates the local production scheduling results based on the theoretical working time, resource consumption, and delivery requirements of the redo process.

[0079] When the quality anomaly handling result is release, the system only updates the quality handling status and / or process status, without triggering production scheduling calculation, performing incremental sorting and rearranging, or updating the incremental production scheduling result.

[0080] By differentiating between rework, redo, and release paths, the system can maintain local update boundaries in scenarios of quality anomalies. For rework or redo paths that require rescheduling of production resources, the system performs local incremental scheduling; for release paths that do not require changes to the scheduling plan, the system only updates the business status, thereby avoiding unnecessary scheduling calculations.

[0081] Example 6: Parameter Adjustment Driven by Approval Status Change Events When the approval status of applications for extra working hours, production schedule adjustments, temporary equipment occupancy, expedited orders, delayed delivery, or quality disposal changes, the system identifies the event as an approval status change event.

[0082] When the target event is an approval status change event, the system dynamically adjusts the available time window parameters of the machine, the theoretical working time parameters of the process, and / or the priority weight of the process based on the approval result. For example, when an application for additional working hours is approved, the system can increase the available processing time window of the corresponding machine or production line within a specified time interval; when an expedited order is approved, the system can increase the order delivery priority of the corresponding order or process; when a production scheduling adjustment application is approved, the system can update the planned time window or resource occupancy constraints of the corresponding process.

[0083] The system further determines whether the corresponding production task is in an incomplete state. If the corresponding production task is in an incomplete state, the system triggers an incremental production schedule update within the corresponding local scheduling domain based on the approval result, and recalculates the theoretical start time, theoretical end time, and incremental production schedule result after conflict resolution within the local scheduling domain. If the corresponding production task is in a completed state, the system only records the approval result for data traceability, without triggering production schedule calculation, performing incremental sorting rearrangement, or updating the incremental production schedule result.

[0084] Through the above processing, approval status change events can be linked with scheduling parameters, while avoiding the pointless rescheduling of completed tasks due to changes in approval status.

[0085] Example 7: Feedback Correction of Deviation Between Theoretical Working Time and Actual Processing Time The system determines the order delivery time window based on the order delivery date, and then deducts periods of resource unavailability and / or production line downtime within the order delivery time window to obtain the effective processing time window. The system proportionally allocates the effective processing time window according to the theoretical time weight of each process to determine the theoretical start time and theoretical end time of each process.

[0086] During the execution of a process, the system collects data such as the actual start time, actual end time, actual processing duration, resources used, operators, processing specifications, and processing status. It then compares the actual processing duration with the theoretical processing time to generate deviation data between the theoretical and actual processing times.

[0087] The system performs statistical analysis on the deviation data based on the process dimension and / or resource dimension, generating time deviation analysis results, correction reference information, and / or candidate parameter adjustment suggestions. The time deviation analysis results include at least one of the following: average deviation, maximum deviation, deviation ratio, or deviation trend of a certain process within a preset time period; the correction reference information includes at least one of the following: suggested theoretical time weight, candidate correction coefficient, or abnormal deviation indication.

[0088] In this embodiment, the proposed candidate parameter adjustments do not automatically overwrite the existing theoretical time parameters of the process. Instead, they are provided for manual confirmation and adjustment of the theoretical time weights and / or the theoretical time parameters of the process. The manually confirmed parameters can be written into the system and used as reference input for subsequent local scheduling domain production.

[0089] Through the above methods, the system can continuously adjust the production scheduling parameters based on actual production data, thereby improving the consistency between subsequent local incremental production scheduling results and actual production conditions.

[0090] Example 8: Pseudo-increment protection and plan stability verification under high-frequency disturbance scenarios In non-standard discrete manufacturing scenarios, the production system continuously operates and handles a variety of high-frequency disturbance events, including new order insertion, equipment malfunctions, process delays, quality rework, manual order insertion, and approval adjustments.

[0091] When multiple target events occur consecutively within a short period, the system classifies and identifies each target event and constructs corresponding local scheduling domains based on the process dependencies, resource conflicts, and time window overlaps of the target events. For processes that do not have process dependencies, resource conflicts, or time window overlaps with each other, the system does not include them in the current local scheduling domain, nor does it participate in the production scheduling calculation for the current scheduling cycle.

[0092] During the scheduling recalculation process, the system limits scheduling recalculation requests and write-back requests through local field identifiers and / or write control fields, and performs consistency checks on the scope of the local recalculation dataset, the scope of the write-back target, and the scheduling call chain. If a scheduling call chain is detected to have a process of performing global production scheduling recalculation first and then filtering local results, the system determines that it does not belong to a true local incremental rescheduling and refuses to execute the corresponding scheduling recalculation request or write-back request, or downgrades it to only updating the business status without triggering production scheduling calculation.

[0093] Through the above processing, the system can maintain the local scheduling boundary under high-frequency disturbance scenarios, so that the update of production scheduling results is limited to the local scheduling domain affected by the target event, avoiding the repeated recalculation of all orders or all processes, reducing meaningless fluctuations in the planned time of unrelated processes, and improving the stability and real-time response capability of production planning.

Claims

1. A real-time production control method based on incremental rearrangement, characterized in that, Includes the following steps: S100: Monitor target events in the manufacturing system and classify and identify the target events; based on the category of the target events, determine whether a scheduling recalculation request is triggered. If a scheduling recalculation request is triggered, proceed to step S200; otherwise, continue monitoring. S200. Identify the set of processes that have process dependencies or resource conflicts with the target event, and construct a local scheduling domain; set a freeze flag and / or write control fields to processes outside the local scheduling domain, so that processes outside the local scheduling domain do not participate in production scheduling calculation in the current scheduling cycle, and their production scheduling records are not updated or rearranged by the production scheduling calculation results of the current scheduling cycle. S300. Based on the local scheduling domain, obtain the data to be recalculated, and extract, generate or filter the local recalculation dataset from the data to be recalculated; the set of processes corresponding to the local recalculation dataset does not exceed the local scheduling domain. S400. Based on the local recalculation dataset, within the local scheduling domain, the effective processing time window is proportionally allocated, and the theoretical start time and theoretical end time of each process are calculated. S500. Based on the local recalculation dataset, detect time conflicts and resource conflicts within the local scheduling domain, and obtain a subset of conflicting procedures. S600. Based on the multi-rule priority model, perform incremental sorting and rearrangement on the conflicting process subset, and update the incremental scheduling result of the local scheduling domain. S700: Write back the updated scheduling results, and limit the write-back scope to the local scheduling domain based on the local domain identifier and / or write control field, so that the scheduling records of processes outside the local scheduling domain are not written, overwritten, updated or rearranged.

2. The real-time production control method based on incremental rearrangement according to claim 1, characterized in that: In step S100, the target event includes at least one of the following: order change event, equipment status change event, process status change event, quality anomaly event, manual intervention event, and approval status change event; After determining the triggering of the scheduling recalculation request in step S100, the scheduling recalculation request is detected; if the scheduling recalculation request points to the global production scheduling recalculation of all orders or all processes, the scheduling recalculation request is intercepted, blocked, or blocked at the entry point; if the scheduling recalculation request does not point to the global production scheduling recalculation of all orders or all processes, then proceed to step S200.

3. The real-time production control method based on incremental rearrangement according to claim 1, characterized in that: In step S200, based on the pre-established and maintained process dependency graph and resource occupation graph, the affected edges or nodes are updated in conjunction with the target event, and within the range of satisfying at least one of the graph traversal depth threshold and resource conflict hop count threshold, the set of processes that have process dependency relationship or resource conflict relationship with the target event is identified.

4. The real-time production control method based on incremental rearrangement according to claim 3, characterized in that: In non-standard discrete manufacturing scenarios where a Bill of Materials (BOM) is missing or the process route changes dynamically, a process sequence and its sequential constraints are generated based on the order-process mapping relationship, the process rule base, and / or the real-time collected work-in-process status data. A process dependency graph is constructed or updated, and the local scheduling domain is constructed based on the updated process dependency graph.

5. The real-time production control method based on incremental rearrangement according to claim 3 or 4, characterized in that: Step S200 further includes limiting the maximum number of affected processes and filtering out non-overlapping processes based on the overlap between the target event time window and the process planning time window, thereby further narrowing the local scheduling domain.

6. The real-time production control method based on incremental rearrangement according to claim 1, characterized in that: In step S300, the data to be recalculated includes at least one of order data, process data, resource usage data, calendar data, and work-in-process status data; the local recalculation dataset is generated by querying based on event-process mapping index, process-resource inverted index, change log incremental cursor, and / or process dependency closure; freezing the scheduling results of processes not included in the local recalculation dataset includes setting a freeze flag for the processes outside the local scheduling domain and / or writing it into the control field.

7. The real-time production control method based on incremental rearrangement according to claim 1, characterized in that: It also includes setting local field identifiers and / or writing control fields for scheduling recalculation requests and write-back requests, and performing consistency checks on the scheduling recalculation requests and write-back requests in subsequent steps; During the scheduling recalculation process triggered by the target event, a local scheduling recalculation channel is established, and global scheduling recalculation requests pointing to all orders or all processes are intercepted, blocked, or blocked at the entry point. If the scheduling recalculation request or write-back request is found to lack a local domain identifier, the local recalculation dataset range exceeds the local scheduling domain, the write-back target includes a process outside the local scheduling domain, or the scheduling call chain has a process of first performing global production scheduling recalculation and then filtering local results, it is determined that the consistency check has failed. When the consistency check fails, the scheduling recalculation request or write-back request is rejected, the global production scheduling recalculation entry is blocked, or the scheduling recalculation request or write-back request is downgraded to only updating the business status without triggering production scheduling calculation.

8. The real-time production control method based on incremental rearrangement according to claim 1, characterized in that: Step S400 includes obtaining an effective processing time window within the local scheduling domain after deducting non-production periods based on the local recalculation dataset, and allocating the effective processing time window proportionally based on the theoretical working time weight of each process, and calculating the theoretical start time and theoretical end time of each process. Step S400 further includes collecting the theoretical working time and actual processing time of the process, and generating working time deviation analysis results, correction reference information and / or candidate parameter adjustment suggestions based on the deviation between the theoretical working time and the actual processing time, and manually confirming and adjusting the theoretical working time weight and the theoretical working time of the process; The theoretical working time weight is used to generate candidate correction coefficients through an error feedback correction model. The error feedback correction model is based on the deviation between theoretical working time and actual processing time according to the process dimension and / or resource dimension, and generates the candidate correction coefficients according to a preset time period and / or according to event triggering. After manual confirmation, the theoretical working time weight and / or process theoretical working time parameters are updated based on the candidate correction coefficients.

9. The real-time production control method based on incremental rearrangement according to claim 1, characterized in that: The multi-rule priority model includes one or more priority rules; the priority rules include at least one or more of the following: order delivery date priority, critical path weight of process, resource bottleneck weight, work-in-process waiting time weight, and manual intervention priority. The incremental sorting and rearrangement includes generating a priority sequence for processes with time conflicts and / or resource conflicts, and based on the priority sequence, performing local adjustments on a subset of conflicting processes using an incremental solution method to resolve the conflicts. The local adjustment includes at least one or more of local swaps, local insertions, and local moves, and the incremental sorting rearrangement is limited to the conflicting subset of operations and its adjacent affected operations.

10. A real-time production control system based on incremental rearrangement, characterized in that, include: The event listening and classification module is used to listen to target events in the manufacturing system, classify and identify the target events, and determine whether to trigger a scheduling recalculation request based on the category of the target event. The local scheduling domain construction module is used to identify the set of processes that have process dependencies or resource conflicts with the target event and construct the local scheduling domain. The local recalculation dataset control module is used to extract, generate, or filter a local recalculation dataset from the data to be recalculated based on the local scheduling domain. The local recalculation dataset control module is also used to set a freeze flag and / or write control fields to processes outside the local scheduling domain, so that processes outside the local scheduling domain do not participate in production scheduling calculation in the current scheduling cycle, and their production scheduling records are not updated or rearranged by the production scheduling calculation results of the current scheduling cycle. The time window ratio calculation module is used to allocate the effective processing time window proportionally within the local scheduling domain based on the local recalculation dataset, and to calculate the theoretical start time and theoretical end time of each process. The conflict detection and incremental sorting module is used to detect time conflicts and resource conflicts in the local scheduling domain based on the local recalculation dataset, obtain a subset of conflicting processes, and perform incremental sorting and rearrangement of the subset of conflicting processes based on a multi-rule priority model to update the incremental scheduling results of the local scheduling domain. The pseudo-incremental protection execution control module is used to set local domain identifiers and / or write control fields for scheduling recalculation requests and write-back requests, and to perform consistency checks on the scheduling recalculation requests and write-back requests. If it is detected that the scheduling recalculation request or write-back request does not carry a local domain identifier, the range of the local recalculation dataset exceeds the local scheduling domain, the write-back target includes a process outside the local scheduling domain, or the scheduling call chain has a process of first performing global production scheduling recalculation and then filtering local results, it is considered that the consistency check has failed. When the consistency check fails, the pseudo-incremental protection execution control module refuses to execute the scheduling recalculation request or write-back request, blocks the global production scheduling recalculation entry, or downgrades the scheduling recalculation request or write-back request to only update the business status without triggering production scheduling calculation. The write-back module is used to write back the updated scheduling results and control the write-back range of the scheduling results based on the local domain identifier and / or write control field, so that the scheduling records of processes outside the local scheduling domain are not written, overwritten, updated or rearranged.

Images