An integrated production collaborative management and control system based on MOM and digital twins

Abstract

The present invention relates to the field of production collaboration technology, and specifically to an integrated production collaboration management and control system based on MOM and digital twins, comprising: a collaborative perception trigger module, which sets a change monitoring mark for each production data at the distributed data acquisition node end to determine whether the production data meets the collaborative perception trigger condition; a fusion compression filtering module, which triggers a data synchronization mechanism to obtain synchronized data if the trigger condition is met; processes the synchronized data based on a multi-level compression strategy to obtain compressed data; a collaborative scheduling and transmission module, which dynamically allocates data packets to the compressed data through a data packet collaborative scheduling model, introduces a "millisecond timestamp-version number" joint verification mechanism, marks the allocated data packets, and transmits the latest data packets to each MOM system; a production data restoration module, which restores the original data structure according to the latest data packet, automatically identifies abnormal fields, compensates for the abnormal fields, and recovers the production data.

Description

Technical Field

[0001] The present invention relates to the field of production collaboration technology, and specifically to an integrated production collaboration management and control system based on MOM and digital twins. Background Art

[0002] In modern manufacturing scenarios, Manufacturing Execution Systems (MES) are evolving into more highly integrated Manufacturing Operations Management (MOM) systems, which coordinate key tasks within the production process, including resource scheduling, task management, and process traceability. Simultaneously, the development of digital twin technology enables state synchronization and simulation between physical entities and information systems, providing a new path for digital modeling and decision support in production processes.

[0003] However, in the collaborative integration of MOM and digital twin systems, the core challenge is no longer just the automation of control processes or management coordination at the organizational level, but how to achieve cross-system data semantic consistency, scheduling information interaction and reasoning support with efficient, dynamic and scalable data structures and processing mechanisms.

[0004] Specifically, the existing system has the following major bottlenecks: First, cross-domain scheduling data processing obstacles. In a distributed production environment, the interaction of scheduling information between different factories or different MOM systems needs to go through standardized data processing and dynamic synchronization mechanisms. However, there is currently a lack of a data scheduling engine that can automatically generate scheduling data dependency graphs, parallel execution sequences, and feedback links based on real-time resource status and task dependencies. Second, the digital twin and MOM data structures are incompatible: Most existing digital twin models are built based on the custom structure of the simulation engine, and lack a unified data abstraction layer for the semantic representation of task units, equipment status, process flows, etc. in the MOM system. This makes it difficult to directly embed digital twins into the scheduling reasoning process, resulting in increased complexity in coupling between systems. Fourth, there is a lack of a reconfigurable data processing framework: Current MOM and digital twin integration systems are mostly proprietary structures, lacking a data processing framework for low-code semantic configuration, making it difficult to dynamically adapt to newly added production units or task flows.

[0005] Therefore, how to provide an integrated system architecture with cross-factory collaborative scheduling capabilities to achieve deep integration of the MOM system and the digital twin system, and meet the intelligent management and control needs in complex collaborative manufacturing scenarios, is an urgent problem to be solved.

[0006] To this end, an integrated production collaborative management and control system based on MOM and digital twins is proposed. Summary of the Invention

[0007] The purpose of the present invention is to provide an integrated production collaborative management and control system based on MOM and digital twins to achieve low-code modeling and semantic-driven generation, while ensuring adaptive evolution and efficient operation in multi-factory collaborative scheduling. The system includes: a collaborative perception trigger module, which sets a change monitoring mark for each production data at the distributed data acquisition node end to determine whether the production data meets the collaborative perception trigger condition; a fusion compression filtering module, which triggers the data synchronization mechanism if the trigger condition is met to obtain synchronized data; processes the synchronized data based on a multi-level compression strategy to obtain compressed data; a collaborative scheduling transmission module, which dynamically allocates data packets to compressed data through a data packet collaborative scheduling model, and introduces a "millisecond timestamp-version number" joint verification mechanism to mark the allocated data packets and transmit the latest data packets to each MOM system; a production data restoration module, which restores the original data structure according to the latest data packet, automatically identifies abnormal fields, and compensates for the abnormal fields to restore production data.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] An integrated production collaborative management and control system based on MOM and digital twins, including:

[0010] A collaborative perception triggering module is used to set a change monitoring mark for each production data at the distributed data collection node end to determine whether the production data meets the collaborative perception triggering conditions;

[0011] A fusion compression and filtering module is used to trigger the data synchronization mechanism and obtain synchronized data if the collaborative perception trigger condition is met; process the synchronized data based on a multi-level compression strategy to obtain compressed data; the multi-level compression strategy includes a first-level pre-compression processing layer, a second-level main compression algorithm selection layer, and a third-level post-processing compression optimization layer;

[0012] The collaborative scheduling and transmission module is used to dynamically allocate data packets to the compressed data according to the current state through the data packet collaborative scheduling model, introduce a "millisecond timestamp-version number" joint verification mechanism, mark the allocated data packets, and transmit the latest data packets to each MOM system;

[0013] The production data restoration module is used to restore the original data structure according to the latest data packet, automatically identify abnormal fields, and compensate for the abnormal fields based on local historical data to restore the production data.

[0014] Preferably, the change monitoring mark includes: a first mark, a second mark and a third mark;

[0015] The first mark is used to determine whether the difference between the current value of the production data and the value at the previous moment exceeds a preset absolute threshold; the second mark is used to calculate whether the rate of change within the sliding time window exceeds a preset dynamic threshold; the third mark obtains a joint state by setting the logical dependency relationship between the data fields, and identifies whether the joint state has any variation behavior that violates the preset rules.

[0016] Preferably, the collaborative sensing triggering condition includes: a first triggering condition, a second triggering condition and a third triggering condition;

[0017] The first trigger condition is that any one of the first mark, the second mark and the third mark in the change monitoring mark is activated; the second trigger condition is that the preset synchronization time interval is reached; the third trigger condition is that the state of key production equipment changes.

[0018] Preferably, the first-level pre-compression processing layer performs classification processing according to the data type of the synchronization data, adopts normalization and quantization pre-processing for numerical data, adopts dictionary coding pre-processing for text data, and adopts trend separation pre-processing for time series data, to obtain first compressed data;

[0019] The second-level main compression algorithm selection layer calculates a compression factor based on a first change value, a second change value, and a third change value; the first change value is a numerical change amplitude of the synchronization data; the second change value is a data change frequency of the synchronization data; and the third change value is a data correlation coefficient of the synchronization data; a compression algorithm is adaptively selected based on the compression factor to compress the first compressed data to obtain second compressed data; the compression algorithm includes: differential encoding with a high compression ratio when the first change value is less than a preset change value, incremental compression when the second change value is higher than a preset frequency, and vector quantization compression when the third change value is higher than a data correlation threshold.

[0020] The third-level post-processing compression optimization layer monitors the bandwidth status, delay status and packet loss rate of the transmission network in real time to evaluate the network quality; dynamically adjusts the compression parameters according to the network quality, improves the compression ratio when the network is congested, and balances the compression ratio and processing speed when the network is idle; and post-processes the second compressed data according to the adjusted compression parameters to obtain compressed data.

[0021] Preferably, the data packet collaborative scheduling model includes: a priority determination unit, a resource status perception unit, a packet structure adjustment unit and a path selection unit;

[0022] The priority determination unit calculates the data packet priority corresponding to the production data based on the business type, data urgency level and equipment status information of the production data; the resource status perception unit identifies the network transmission status based on the network quality and the predicted congestion level; the packet structure adjustment unit dynamically allocates data packets by adaptively adjusting the granularity, field redundancy rate and fault-tolerant field ratio of a single data packet based on the data packet priority and the network transmission status; the path selection unit selects the optimal transmission path based on the data packet priority, the target MOM system location and the network transmission status, combined with the shortest path first algorithm and the traffic engineering algorithm, and performs multi-path parallel transmission on the allocated data packets through the optimal transmission path.

[0023] Preferably, the process of marking the allocated data packets is:

[0024] Millisecond-level timestamp information is attached to each data packet, and a version number identifier is generated synchronously; the version number is generated based on the historical evolution path and compression rounds of the production data; a "millisecond-level timestamp-version number" joint verification mechanism is introduced to perform a double verification of the time sequence consistency and version integrity of the currently allocated data packet at the receiving end, and the latest data packet with the latest timestamp and version information is selected and transmitted to each MOM system.

[0025] Preferably, the automatic identification of abnormal fields and compensation for the abnormal fields based on local historical data are specifically carried out as follows:

[0026] Compare the latest data field value of the latest data packet received in real time with local historical data to preliminarily mark potential abnormal fields; call the digital twin model, input the context data of the potential abnormal field, and obtain the theoretical predicted value of the potential abnormal field under the current working conditions; if the difference between the latest data field value and the theoretical predicted value exceeds a preset difference threshold, it is confirmed as an abnormal field;

[0027] Based on the historical trends and current values ​​of the abnormal fields and normal fields, a compensation value is generated; if the confidence of the model compensation value is lower than a preset confidence threshold, the compensation value is corrected using the process knowledge and / or physical constraints encapsulated in the digital twin model; and the production data is restored based on the compensation value.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] 1. The present invention realizes real-time status perception and intelligent synchronization triggering of production data by setting multi-dimensional change monitoring marks. Compared with the traditional synchronization mechanism that relies on fixed time intervals or single field judgments, the present invention introduces field difference thresholds, change rate thresholds and logical dependency judgments between fields, which can more sensitively capture meaningful data changes. The change monitoring mark effectively reduces the frequency of invalid synchronization, improves the response speed to key data changes, significantly enhances the timeliness and accuracy of data acquisition by the MOM system, and provides stable and reliable original input data for the digital twin model.

[0030] 2. The present invention proposes a multi-level compression strategy that performs pre-processing compression based on data types and characteristics, dynamically selects a primary compression algorithm based on change values, frequency, and correlation indicators, and performs post-processing optimization based on the real-time network status, thereby achieving intelligent compression scheduling of transmitted data. This multi-level compression strategy not only improves the system's data transmission efficiency, but also enhances its adaptability to different network environments, effectively alleviating transmission bottlenecks and ensuring stable and efficient transmission of key production data even under network congestion conditions, providing data security for the continuous operation of the MOM system.

[0031] 3. This invention constructs an abnormal field identification and compensation method based on the digital twin model. By jointly analyzing data context, historical trends, and operating condition predictions, it effectively identifies data anomalies and implements intelligent repair. Compared with traditional mean filling or static regression methods, this method can dynamically generate high-confidence compensation results and further verify and correct them by combining built-in process logic and physical rules. This significantly improves the accuracy and reliability of data recovery, provides higher-quality data support for subsequent production control and intelligent decision-making, and ensures adaptive evolution and efficient operation in multi-factory collaborative scheduling. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 A schematic diagram of the structure of an integrated production collaborative management and control system based on MOM and digital twins provided in an embodiment of the present invention;

[0033] Figure 2 A schematic diagram of the integrated production collaborative management and control process based on MOM and digital twins provided in an embodiment of the present invention;

[0034] Figure 3 A schematic diagram of a process for triggering a data synchronization mechanism according to an embodiment of the present invention;

[0035] Figure 4 A diagram illustrating the working principle of the multi-stage compression strategy provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0036] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0037] The present invention proposes an integrated production collaborative management and control system based on MOM and digital twins, which can achieve low-code modeling and semantic-driven generation, while ensuring adaptive evolution and efficient operation in multi-factory collaborative scheduling. In order to illustrate that the method of the present invention can achieve low-code modeling and semantic-driven generation, while ensuring adaptive evolution and efficient operation in multi-factory collaborative scheduling, the effectiveness of the present invention will be illustrated from two examples below.

[0038] Example 1:

[0039] This embodiment is applied to the multi-factory collaborative production environment of a large manufacturing enterprise, which has production bases in regions A and B and requires cross-regional collaborative production management and control. The integrated production collaborative management and control system of the present invention is integrated with multiple MOM subsystems into the enterprise's digital platform. The core technical architecture of this digital platform is a multi-layered platform, spanning automation, informationization, digitization, and ultimately intelligence. The bottom layer comprises the equipment and control layers, encompassing various automated production lines, sensors, PLCs, DCSs, and more. The execution layer, the core of the present invention, is built on a MOM (Manufacturing Operations Management) platform that integrates multiple functional modules, including data acquisition SCADA, logistics and warehousing (WMS), factory operations management (FMCS), equipment management (EAM), energy management (EMS), intelligent manufacturing (MES), and digital twin (DTwin). On top of the MOM platform, management manages the supply chain and value chain through systems such as ERP, SRM, and CRM. Ultimately, the decision-making layer utilizes AI big models and BI tools to analyze the enterprise's monthly, quarterly, and annual performance data to support decision-making.

[0040] In an embodiment of the present application, the system proposed in the present invention is used to describe in detail the process of achieving low-code modeling and semantic-driven generation while ensuring adaptive evolution and efficient operation in multi-factory collaborative scheduling. Figure 1This is a specific structural diagram of the system of the present invention, including: a collaborative perception trigger module, which sets a change monitoring mark for each production data at the distributed data acquisition node end to determine whether the production data meets the collaborative perception trigger condition; a fusion compression filtering module, which triggers the data synchronization mechanism if the trigger condition is met to obtain synchronized data; processes the synchronized data based on a multi-level compression strategy to obtain compressed data; a collaborative scheduling and transmission module, which dynamically allocates data packets to compressed data through a data packet collaborative scheduling model, and introduces a "millisecond timestamp-version number" joint verification mechanism to mark the allocated data packets and transmit the latest data packets to each MOM system; a production data restoration module, which restores the original data structure according to the latest data packet, automatically identifies abnormal fields, compensates for the abnormal fields, and recovers production data. Figure 2 The following is a flow chart of the integrated production collaborative management and control based on MOM and digital twins provided by the embodiment of the present invention. Figure 1 and Figure 2 The following content is described:

[0041] An integrated production collaborative management and control system based on MOM and digital twins, including:

[0042] A collaborative perception triggering module is used to set a change monitoring mark for each production data at the distributed data collection node end to determine whether the production data meets the collaborative perception triggering conditions;

[0043] Preferably, in the injection molding workshop of the factory in area A, its production operation management activities are managed by a set of intelligent manufacturing execution systems (ForceCon-MES) to set three change monitoring marks for the production data of key equipment such as injection molding machines: the change monitoring marks include: a first mark, a second mark, and a third mark;

[0044] The first mark is used to determine whether the difference between the current value of the production data and the value at the previous moment exceeds a preset absolute threshold; for example, when a significant change in the mold temperature of the injection molding machine is detected, the temperature difference value exceeds the preset absolute threshold, and the first mark is activated.

[0045] The second mark is used to calculate whether the rate of change within the sliding time window exceeds a preset dynamic threshold; for example, when it is detected that the production rhythm of the injection molding machine fluctuates greatly within the set time window, the rate of change exceeds the preset dynamic threshold, and the second mark is activated.

[0046] The third flag, which establishes a logical dependency between data fields to generate a joint state, identifies whether the joint state exhibits variations that violate predefined rules. For example, if the temperature rises but the pressure drops abnormally, the joint state exhibits variations that violate predefined rules, and the third flag is activated.

[0047] The collaborative perception and triggering module of this embodiment provides three change monitoring markers, enabling more refined, multi-dimensional, and intelligent monitoring and identification of production data changes. These markers allow the system to more accurately determine whether meaningful data changes have occurred, avoiding unnecessary synchronization and improving the pertinence and effectiveness of synchronization. This in turn increases the speed and sensitivity of response to key data changes, enabling real-time status awareness of production data and intelligent synchronization triggering.

[0048] Preferably, reference Figure 3 , the collaborative sensing triggering conditions include: a first triggering condition, a second triggering condition and a third triggering condition;

[0049] The first trigger condition is that any one of the first mark, the second mark and the third mark in the change monitoring mark is activated;

[0050] The second trigger condition is reaching a preset synchronization time interval;

[0051] The third trigger condition is a change in the status of key production equipment; for example, when the key equipment of Factory A changes from the "running" status to the "fault" status, data synchronization is triggered immediately. This status information will not only update the digital twin model, but will also be displayed in real time on the electronic dashboard in the workshop, and notify other factories to adjust their production plans.

[0052] This embodiment provides three specific conditions for triggering data synchronization, providing a flexible and comprehensive data synchronization triggering mechanism, ensuring the timeliness and necessity of data synchronization. This multi-condition combination triggering method enables the system to respond to immediate and significant data changes, meet periodic update requirements, and quickly synchronize key events (such as device status changes), thereby ensuring the timeliness and accuracy of data acquired by the MOM system.

[0053] Preferably, the fusion compression filtering module is used to trigger the data synchronization mechanism to obtain synchronized data if the collaborative perception triggering condition is met; the synchronized data is processed based on a multi-level compression strategy to obtain compressed data; the multi-level compression strategy includes a first-level pre-compression processing layer, a second-level main compression algorithm selection layer and a third-level post-processing compression optimization layer; reference Figure 4 ;

[0054] The first-level pre-compression processing layer performs classification processing according to the data type of the synchronization data, adopts normalization and quantization pre-processing for numerical data, adopts dictionary coding pre-processing for text data, and adopts trend separation pre-processing for time series data, to obtain first compressed data;

[0055] The second-level main compression algorithm selection layer adaptively selects a compression algorithm based on a compression factor to compress the first compressed data to obtain second compressed data; the compression factor includes a first change value, a second change value, and a third change value; the first change value is the numerical change amplitude of the synchronization data; the second change value is the data change frequency of the synchronization data; and the third change value is the data correlation coefficient of the synchronization data; the compression algorithm includes: when the first change value is less than a preset change value, differential encoding with a high compression ratio is used; when the second change value is higher than a preset frequency, incremental compression is used; and when the third change value is higher than a data correlation threshold, vector quantization compression is used.

[0056] The third-level post-processing compression optimization layer monitors the bandwidth status, delay status and packet loss rate of the transmission network in real time to evaluate the network quality; dynamically adjusts the compression parameters according to the network quality, improves the compression ratio when the network is congested, and balances the compression ratio and processing speed when the network is idle; and post-processes the second compressed data according to the adjusted compression parameters to obtain compressed data.

[0057] This embodiment details the multi-stage compression strategy employed by the fused compression and filtering module, including a first-stage pre-compression process tailored to different data types, a second-stage process that adaptively selects the primary compression algorithm based on data variation characteristics, and a third-stage post-processing process that dynamically adjusts compression parameters based on network status. This achieves efficient, intelligent data compression that adapts to changing network environments, significantly improving data transmission efficiency and enhancing network adaptability. Pre-processing reduces redundancy, adaptive algorithm selection improves compression ratio and efficiency, and network-aware post-processing ensures transmission performance under varying network conditions, effectively alleviating transmission bottlenecks.

[0058] Preferably, the collaborative scheduling transmission module is used to dynamically allocate data packets to the compressed data through the data packet collaborative scheduling model according to the current state, and introduce a "millisecond timestamp-version number" joint verification mechanism to mark the allocated data packets and transmit the latest data packets to each MOM system;

[0059] The data packet collaborative scheduling model includes: a priority determination unit, a resource status perception unit, a packet structure adjustment unit and a path selection unit;

[0060] The priority determination unit calculates the data packet priority corresponding to the production data based on the business type, data urgency level and equipment status information of the production data; the resource status perception unit identifies the network transmission status based on the network quality and the predicted congestion level; the packet structure adjustment unit dynamically allocates data packets by adaptively adjusting the granularity, field redundancy rate and fault-tolerant field ratio of a single data packet based on the data packet priority and the network transmission status; the path selection unit selects the optimal transmission path based on the data packet priority, the target MOM system location and the network transmission status, combined with the shortest path first algorithm and the traffic engineering algorithm, and performs multi-path parallel transmission on the allocated data packets through the optimal transmission path.

[0061] Specifically, the priority determination unit classifies the injection molding machine fault warning data as the "equipment monitoring" business type, sets the urgency level to "high", and the equipment status to "abnormal", and calculates a higher data packet priority.

[0062] Resource status perception unit: predicts the network transmission status from A to B, combines historical data and current load conditions, and identifies the network transmission status.

[0063] Packet structure adjustment unit: Adaptively adjusts packet parameters for high-priority packets and network congestion to improve transmission reliability.

[0064] Path selection unit: selects the optimal transmission path from area A → Q node → P node → area B, and enables backup paths for multi-path parallel transmission.

[0065] This embodiment proposes a collaborative data packet scheduling model encompassing four components: priority determination, resource status awareness, packet structure adjustment, and path selection. This model enables intelligent and refined allocation of compressed data packets and efficient, reliable multi-path parallel transmission. By sensing data priority and network status, it dynamically adjusts packet structure and selects the optimal transmission path, ensuring prioritized, efficient, and stable transmission of important data to the target MOM system. This optimizes resource utilization and ensures transmission quality, particularly in complex network environments.

[0066] Preferably, the process of marking the allocated data packets is:

[0067] Millisecond-level timestamp information is attached to each data packet, and a version number identifier is generated synchronously; the version number is generated based on the historical evolution path and compression rounds of the production data; a "millisecond-level timestamp-version number" joint verification mechanism is introduced to perform a double verification of the time sequence consistency and version integrity of the currently allocated data packet at the receiving end, and the latest data packet with the latest timestamp and version information is selected and transmitted to each MOM system.

[0068] Specifically, the millisecond timestamp-version number joint verification mechanism:

[0069] Each data packet is appended with a millisecond timestamp, such as "2024-06-04 14:23:45.872," accurate to the millisecond level. A version number is also generated simultaneously. The version number is based on the historical evolution path of the production data and the number of compression rounds, in the format of "V2024.06.04.001.C3," where C3 indicates the third round of compression.

[0070] At the receiving end, the time sequence consistency and version integrity of the currently allocated data packets are double-checked. Upon receiving the data packets from Factory A, Factory B's MOM system first checks the timestamp sequence to ensure that the timestamps of later arriving packets are no earlier than those of earlier ones. Secondly, it verifies the integrity of the version numbers to ensure that the sequence is continuous and complete. The latest data packets with the latest timestamps and version information are selected and transmitted to each MOM system.

[0071] This embodiment specifically describes the process of marking and verifying allocated data packets using a combined millisecond timestamp-version number verification mechanism. This process ensures the time sequence consistency and version integrity of transmitted data packets at the receiving end, thereby ensuring that each MOM system receives and uses the latest, correct data. This dual verification effectively avoids data disarray or the use of outdated data, providing an accurate data foundation for subsequent data recovery and application. Finally, only the latest version of the data packet is transmitted, effectively avoiding invalid transmissions and redundant resource waste, thereby improving transmission speeds.

[0072] Preferably, the production data restoration module is used to restore the original data structure according to the latest data packet, automatically identify abnormal fields, and compensate for the abnormal fields based on local historical data to restore the production data; the specific process is:

[0073] After receiving the latest data packet, Factory B compares the latest data field values ​​of the latest data packet received in real time with the local historical data and preliminarily marks potential abnormal fields; for example, when the injection molding machine temperature data deviates significantly from the historical average, it is marked as a potential abnormal field.

[0074] Calling the digital twin model, inputting the context data of the potential abnormal field, and obtaining the theoretical predicted value of the potential abnormal field under the current working conditions; if the difference between the latest data field value and the theoretical predicted value exceeds a preset difference threshold, it is confirmed as an abnormal field;

[0075] Based on the historical trends and current values ​​of the abnormal and normal fields, a compensation value is generated. If the confidence level of the model compensation value is lower than a preset confidence threshold, the compensation value is corrected using the process knowledge and / or physical constraints encapsulated in the digital twin model. For example, when the confidence level of the preliminary compensation value is insufficient, the system uses the injection molding process knowledge and physical constraints encapsulated in the digital twin model to make corrections, including process parameters such as the optimal processing temperature range of the material and the ideal temperature range corresponding to the current load.

[0076] The production data is restored based on the compensation value to ensure data accuracy and reliability. The restored high-quality data will serve the ForceCon-EMS energy management platform for accurate energy consumption statistics or accurate quality traceability.

[0077] This embodiment defines the specific process for automatically identifying and compensating for abnormal fields in the production data restoration module. It specifically emphasizes using the digital twin model to obtain theoretical predictions to confirm anomalies, and utilizing the process knowledge and physical constraints encapsulated in the digital twin model to make corrections when the compensation values ​​are insufficiently confident. This achieves high-precision identification of abnormal data and highly reliable intelligent compensation, significantly improving the accuracy and reliability of data recovery. By combining historical data, digital twin predictions, and domain knowledge, data anomalies can be more effectively handled, providing high-quality data support for subsequent production control and intelligent decision-making.

[0078] This embodiment provides an integrated, end-to-end data processing flow, aiming to achieve efficient, intelligent, and reliable production collaborative management and control between the manufacturing operations management system (MOM) and the digital twin.

[0079] The collaborative perception trigger module shifts from passive, timed synchronization to proactive, intelligent triggering by setting change monitoring flags for each piece of production data. This mechanism accurately captures meaningful data changes and instantly triggers synchronization, reducing critical data response time from seconds to milliseconds and significantly reducing system resource consumption.

[0080] The fusion compression and filtering module adopts a three-level progressive compression strategy, dynamically selects the optimal compression algorithm according to data characteristics and network status, achieves a high compression ratio, significantly reduces the network transmission load, and has strong adaptability to network environments, ensuring stable transmission performance under different network conditions.

[0081] The collaborative scheduling transmission module introduces an intelligent allocation model and a millisecond-level timestamp-version number verification mechanism to achieve dynamic load balancing and avoid network congestion. Millisecond-level timestamps ensure the consistency of data packet time sequence, and version number verification prevents data loss and duplication, improving data transmission reliability.

[0082] The production data restoration module significantly improves data quality assurance through automatic anomaly identification and intelligent compensation mechanisms. This module intelligently identifies transmission anomalies and accurately compensates for them based on historical data, improving data accuracy and anomaly repair capabilities.

[0083] Through the collaborative work of these four modules, the system ensures full-chain optimization in a distributed environment, from data acquisition, processing, transmission, to final recovery application. This provides a foundational framework for low-code configuration, semantically driven generation, and adaptive evolution and efficient operation in multi-factory collaborative scheduling. This framework supports the deep integration and efficient collaboration of multiple industrial applications on the company's MOM platform, including logistics and warehousing (WMS), facility management (FMCS), equipment management (EAM), energy management (EMS), and intelligent manufacturing (MES).

[0084] Example 2:

[0085] In Example 1, the method proposed in this invention successfully achieved low-code modeling and semantically driven generation, while ensuring adaptive evolution and efficient operation in multi-factory collaborative scheduling. To further verify the effectiveness of this invention, the example of this application also implemented collaborative management and control of cross-factory production processes.

[0086] A collaborative perception triggering module is used to set a change monitoring mark for each production data at the distributed data collection node end to determine whether the production data meets the collaborative perception triggering conditions;

[0087] The change monitoring mark includes: a first mark, a second mark and a third mark;

[0088] The first mark is used to determine whether the difference between the current value of the production data and the value at the previous moment exceeds a preset absolute threshold; the second mark is used to calculate whether the rate of change within the sliding time window exceeds a preset dynamic threshold; the third mark obtains a joint state by setting the logical dependency relationship between the data fields, and identifies whether the joint state has any variation behavior that violates the preset rules.

[0089] The collaborative sensing triggering conditions include: a first triggering condition, a second triggering condition and a third triggering condition;

[0090] The first trigger condition is that any one of the first mark, the second mark and the third mark in the change monitoring mark is activated; the second trigger condition is that the preset synchronization time interval is reached; the third trigger condition is that the state of key production equipment changes.

[0091] The present invention realizes multi-dimensional and in-depth perception of the status of production data by combining the application of change monitoring marks based on absolute thresholds, dynamic change rates and logical dependencies. The setting of this change monitoring mark is far superior to the traditional synchronization method that only relies on fixed time intervals or simple judgments of single fields. It can more sensitively capture real and meaningful data changes, such as changes in the production bottleneck data of a factory, or the energy consumption index of a key equipment exceeds the normal fluctuation range. This intelligent synchronization trigger mechanism effectively reduces the network load and computing resource consumption caused by unnecessary data synchronization, and significantly improves the response speed to changes in key data. In the scenario of multi-factory collaborative scheduling, this means that each MOM system can obtain the effective status updates of other collaborative units more quickly, providing stable and reliable original input data for the real-time and accuracy of the digital twin model.

[0092] Preferably, a fusion compression and filtering module is used to trigger a data synchronization mechanism to obtain synchronized data if a collaborative perception triggering condition is met; the synchronized data is processed based on a multi-level compression strategy to obtain compressed data; the multi-level compression strategy includes a first-level pre-compression processing layer, a second-level main compression algorithm selection layer, and a third-level post-processing compression optimization layer;

[0093] The first-level pre-compression processing layer performs classification processing according to the data type of the synchronization data, adopts normalization and quantization pre-processing for numerical data, adopts dictionary coding pre-processing for text data, and adopts trend separation pre-processing for time series data, to obtain first compressed data;

[0094] The second-level main compression algorithm selection layer calculates a compression factor based on a first change value, a second change value, and a third change value; the first change value is a numerical change amplitude of the synchronization data; the second change value is a data change frequency of the synchronization data; and the third change value is a data correlation coefficient of the synchronization data; a compression algorithm is adaptively selected based on the compression factor to compress the first compressed data to obtain second compressed data; the compression algorithm includes: differential encoding with a high compression ratio when the first change value is less than a preset change value, incremental compression when the second change value is higher than a preset frequency, and vector quantization compression when the third change value is higher than a data correlation threshold.

[0095] The third-level post-processing compression optimization layer monitors the bandwidth status, delay status and packet loss rate of the transmission network in real time to evaluate the network quality; dynamically adjusts the compression parameters according to the network quality, improves the compression ratio when the network is congested, and balances the compression ratio and processing speed when the network is idle; and post-processes the second compressed data according to the adjusted compression parameters to obtain compressed data.

[0096] The multi-level compression strategy proposed in the present invention realizes highly intelligent compression scheduling of transmitted data through first-level preprocessing, dynamic selection of the second-level main compression algorithm, and third-level post-processing optimization based on real-time network status. It is not just a simple application of a general compression algorithm, but a targeted and adaptive compression based on the specific type of data, change characteristics and actual transmission network conditions. This strategy significantly improves the data transmission efficiency and robustness of the system in various network environments. In multi-factory collaborative manufacturing, the network conditions between factories may vary greatly and change dynamically. The multi-level compression strategy can effectively alleviate the information delay caused by data transmission bottlenecks, ensure that key production data, such as urgent order changes, equipment failure alarms, etc., can be stably and efficiently transmitted to the relevant MOM systems even under network congestion conditions, and provide solid data protection for the continuous and efficient operation of the entire collaborative production network.

[0097] Preferably, the collaborative scheduling transmission module is used to dynamically allocate data packets to the compressed data through the data packet collaborative scheduling model according to the current state, and introduce a "millisecond timestamp-version number" joint verification mechanism to mark the allocated data packets and transmit the latest data packets to each MOM system;

[0098] The data packet collaborative scheduling model includes: a priority determination unit, a resource status perception unit, a packet structure adjustment unit and a path selection unit;

[0099] The priority determination unit calculates the data packet priority corresponding to the production data based on the business type, data urgency level and equipment status information of the production data; the resource status perception unit identifies the network transmission status based on the network quality and the predicted congestion level; the packet structure adjustment unit dynamically allocates data packets by adaptively adjusting the granularity, field redundancy rate and fault-tolerant field ratio of a single data packet based on the data packet priority and the network transmission status; the path selection unit selects the optimal transmission path based on the data packet priority, the target MOM system location and the network transmission status, combined with the shortest path first algorithm and the traffic engineering algorithm, and performs multi-path parallel transmission on the allocated data packets through the optimal transmission path.

[0100] Preferably, the process of marking the allocated data packets is:

[0101] Millisecond-level timestamp information is attached to each data packet, and a version number identifier is generated synchronously; the version number is generated based on the historical evolution path and compression rounds of the production data; a "millisecond-level timestamp-version number" joint verification mechanism is introduced to perform a double verification of the time sequence consistency and version integrity of the currently allocated data packet at the receiving end, and the latest data packet with the latest timestamp and version information is selected and transmitted to each MOM system.

[0102] Preferably, the production data restoration module is used to restore the original data structure according to the latest data packet, automatically identify abnormal fields, and compensate for the abnormal fields based on local historical data to restore the production data.

[0103] The automatic identification of abnormal fields and compensation of abnormal fields based on local historical data are as follows:

[0104] Compare the latest data field value of the latest data packet received in real time with local historical data to preliminarily mark potential abnormal fields; call the digital twin model, input the context data of the potential abnormal field, and obtain the theoretical predicted value of the potential abnormal field under the current working conditions; if the difference between the latest data field value and the theoretical predicted value exceeds a preset difference threshold, it is confirmed as an abnormal field;

[0105] Based on the historical trends and current values ​​of the abnormal fields and normal fields, a compensation value is generated; if the confidence of the model compensation value is lower than a preset confidence threshold, the compensation value is corrected using the process knowledge and / or physical constraints encapsulated in the digital twin model; and the production data is restored based on the compensation value.

[0106] This invention develops an advanced method for identifying and compensating abnormal fields. By comparing real-time data with the theoretical predictions of a digital twin model under current operating conditions, it can more accurately identify hidden anomalies that are difficult to detect using traditional methods. When compensation for abnormal data is required, it not only relies on simple historical data statistical methods (such as mean filling and simple regression), but also combines the historical trends and current values ​​of the abnormal field and related normal fields for preliminary compensation. If the confidence level of the compensation value is insufficient, the compensation result is verified and corrected by leveraging the deep process knowledge and physical constraints embedded in the digital twin model. This abnormal field identification and compensation method ensures that the compensation value generation considers both the statistical characteristics of the data and the actual physical and process logic, significantly improving the accuracy, robustness, and reliability of data recovery. In multi-factory collaborative scheduling, high-quality data is the cornerstone of precise control, intelligent decision-making, and system optimization. This method provides purer and more reliable data support for each MOM system and upper-level digital twin applications, ensuring that the entire collaborative manufacturing system can maintain adaptive evolution and efficient operation in a dynamically changing environment.

[0107] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A production collaborative management and control integrated system based on MOM and digital twin, characterized by: include: The data synchronization trigger module is used to set a change monitoring mark for each target data at the distributed data acquisition node end to determine whether the target data meets the data synchronization trigger condition; The data compression and filtering module is used to trigger the data synchronization mechanism and obtain synchronized data if the data synchronization triggering conditions are met; Processing the synchronized data based on a multi-level compression strategy to obtain compressed data; The multi-stage compression strategy includes a first-stage pre-compression processing layer, a second-stage main compression algorithm selection layer, and a third-stage post-processing compression optimization layer; The first-level pre-compression processing layer performs classification processing according to the data type of the synchronization data, adopts normalization and quantization pre-processing for numerical data, adopts dictionary coding pre-processing for text data, and adopts trend separation pre-processing for time series data, to obtain first compressed data; The second-level main compression algorithm selection layer calculates a compression factor based on the first change value, the second change value, and the third change value; The first change value is the magnitude of the numerical change of the synchronization data; The second change value is the data change frequency of the synchronization data; the third change value is the data correlation coefficient of the synchronization data; adaptively selecting a compression algorithm according to the compression factor, compressing the first compressed data to obtain second compressed data; The compression algorithm includes: differential encoding when the first change value is less than a preset change value, incremental compression when the second change value is higher than a preset frequency, and vector quantization compression when the third change value is higher than a data correlation threshold; the third-level post-processing compression optimization layer monitors the bandwidth status, delay status, and packet loss rate of the transmission network in real time to evaluate network quality; dynamically adjusts compression parameters based on the network quality, and post-processes the second compressed data based on the adjusted compression parameters to obtain compressed data; The data allocation and transmission module is used to dynamically allocate data packets to the compressed data according to the current status through the data packet intelligent allocation model, introduce a "millisecond timestamp-version number" joint verification mechanism, mark the allocated data packets, and transmit the latest data packets to each MOM system; The data recovery module is used to restore the original data structure according to the latest data packet, automatically identify abnormal fields, and compensate for the abnormal fields based on local historical data to restore the target data.

2. A production collaborative management and control integrated system based on MOM and digital twin according to claim 1, characterized in that: The change monitoring mark includes: a first mark, a second mark and a third mark; The first mark is used to determine whether the difference between the current value of the target data field and the value at the previous moment exceeds a preset absolute threshold; the second mark is used to calculate whether the rate of change within the sliding time window exceeds a preset dynamic threshold; the third mark obtains a joint state by setting a logical dependency relationship between data fields, and identifies whether the joint state has any variation behavior that violates the preset rules.

3. The integrated production collaborative management and control system based on MOM and digital twin according to claim 1 is characterized in that: The data synchronization triggering conditions include: a first triggering condition, a second triggering condition and a third triggering condition; The first trigger condition is that any one of the first mark, the second mark and the third mark in the change monitoring mark is activated; the second trigger condition is that the preset synchronization time interval is reached; the third trigger condition is that the state of key production equipment changes.

4. The integrated production collaborative management and control system based on MOM and digital twin according to claim 1 is characterized in that: The data packet intelligent distribution model includes: a priority determination unit, a resource status perception unit, a packet structure adjustment unit and a path selection unit; The priority determination unit calculates the data packet priority corresponding to the target data based on the business type, data urgency level and device status information of the target data; the resource status perception unit identifies the network transmission status based on the network quality and the predicted congestion level; the packet structure adjustment unit dynamically allocates data packets by adaptively adjusting the granularity, field redundancy rate and fault-tolerant field ratio of a single data packet based on the data packet priority and the network transmission status; the path selection unit selects the optimal transmission path based on the data packet priority, the target MOM system location and the network transmission status, combined with the shortest path first algorithm and the traffic engineering algorithm, and performs multi-path parallel transmission on the allocated data packets through the optimal transmission path.

5. The integrated production collaborative management and control system based on MOM and digital twin according to claim 1 is characterized in that: The process of marking the allocated data packets is as follows: Millisecond-level timestamp information is attached to each data packet, and a version number is generated synchronously. The version number is generated based on the historical evolution path and compression rounds of the target data field. A "millisecond-level timestamp-version number" joint verification mechanism is introduced to double-check the time sequence consistency and version integrity of the currently allocated data packets at the receiving end, and select the latest data packets with the latest timestamp and version information for transmission to each MOM system.

6. The integrated production collaborative management and control system based on MOM and digital twin according to claim 1 is characterized in that: The automatic identification of abnormal fields and compensation of abnormal fields based on local historical data are as follows: Compare the latest data field value of the latest data packet received in real time with the local historical data to preliminarily mark the potential abnormal field; call the digital twin model, input the context data of the potential abnormal field, and obtain the theoretical prediction value of the potential abnormal field under the current working conditions; If the difference between the latest data field value and the theoretical prediction value exceeds a preset difference threshold, it is confirmed as an abnormal field; Based on the historical trends and current values ​​of the abnormal fields and normal fields, a compensation value is generated; if the confidence of the model compensation value is lower than a preset confidence threshold, the compensation value is corrected using the process knowledge and / or physical constraints encapsulated in the digital twin model; and the target data is restored based on the compensation value.

Images