Modular reconfigurable production line control system integration method

Abstract

The application discloses a kind of modular reconfigurable production line control system integration method, it is related to industrial automation technical field, establishes virtual mapping based on physical attribute parameter, forms production line digital twin basic model framework;Based on real-time data acquisition of field sensor, establishes the dynamic mapping mechanism of digital twin body synchronous with physical production line state;Reconfiguration scheme is imported in virtual environment, the combination process is previewed by production line digital twin model, and key performance indicators are analyzed.The present application constructs virtual production line model through digital twin technology, supports the reconfiguration scheme to be previewed and verified in virtual environment, significantly reduces the trial-and-error time and cost required by traditional physical debugging, cloud AI algorithm further optimizes reconfiguration strategy, combined with edge real-time control, realizes closed-loop optimization from virtual verification to physical execution, so that production line can quickly complete local or overall reconfiguration according to production demand, greatly improves the response speed and flexibility of production line.

Description

Technical Field

[0001] This invention relates to the field of industrial automation technology, and more specifically to a modular and reconfigurable production line control system integration method. Background Technology

[0002] With the rapid changes in market demand, modern manufacturing is increasingly inclined towards customized production. Traditional production lines usually face high equipment fixity and system complexity. However, the design concept of modular and reconfigurable production line control systems allows production lines to be quickly adjusted and reconfigured according to changes in product demand. This flexibility can cope with the production cycles and output requirements of different products, improving the adaptability of the production line.

[0003] In existing technologies, reconfiguration schemes rely on on-site physical adjustments and human experience, lacking virtual simulation verification. Repeated trial and error are required to verify the feasibility and efficiency of module combinations, which is time-consuming and prone to physical debugging failures. Therefore, the problem to be solved by this invention is to construct a digital twin model of the production line, map the physical state in real time, and pre-run the reconfiguration scheme in a virtual environment, verify its feasibility through simulation, and realize global intelligent control by utilizing edge-cloud collaboration technology. To this end, a modular reconfigurable production line control system integration method is proposed. Summary of the Invention

[0004] The purpose of this invention is to provide a modular and reconfigurable production line control system integration method to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0006] A modular and reconfigurable production line control system integration method includes the following steps:

[0007] S1. Establish a virtual mapping based on physical property parameters to form a basic model framework for the production line digital twin;

[0008] S2. Based on real-time data acquisition from on-site sensors, establish a dynamic mapping mechanism for a digital twin that is synchronized with the physical production line status;

[0009] S3. Import the reconstruction scheme into the virtual environment, simulate the module combination process through the production line digital twin model, analyze key performance indicators, and verify the feasibility and efficiency through simulation.

[0010] S4. Optimize the reconstruction plan based on the simulation results, and iteratively update the production line digital twin model until the feasibility and efficiency requirements are met.

[0011] S5. Deploy a lightweight digital twin agent at the edge to perform module-level real-time data collection, status monitoring, and localized control command issuance;

[0012] S6. Build a global digital twin platform in the cloud, integrate data from multiple production lines, use AI algorithms to analyze and generate the optimal reconstruction strategy and distribute it.

[0013] S7: Through real-time interaction between the edge and cloud, control commands are executed at the edge, and strategies are dynamically adjusted in the cloud to complete production line reconstruction and intelligent control.

[0014] A further improvement to the technical solution of this invention lies in the following: In step S1, the formation process of the basic model framework for the production line digital twin is as follows:

[0015] Comprehensive collection of physical property parameters of each module of the production line, covering dimensions, mass, range of motion and dynamic characteristics, etc. The collected parameters are systematically classified and organized, and structured storage is carried out using a unified data standard to ensure data integrity and accuracy.

[0016] Based on the organized physical property parameters, and using parametric modeling tools, virtual models of each module are constructed according to the actual structure and module connection relationships of the production line. These models are then assembled and integrated to form a preliminary virtual model architecture for the production line.

[0017] The initial virtual production line model architecture is verified to check its matching degree with the actual production line in terms of structure and function. Based on the verification results, corrections and adjustments are made to form a basic digital twin model framework for the production line that reflects the characteristics of the production line.

[0018] A further improvement to the technical solution of this invention lies in the following: In step S2, the process of establishing a dynamic mapping mechanism for a digital twin synchronized with the physical production line status is as follows:

[0019] Deploy industrial protocol gateways on the physical production line to connect various field sensors to PLCs and robot controllers, establish data acquisition channels, configure a time series database to store status data output by field sensors, including position, speed, and temperature, and define a data tagging system to give each parameter a unique identifier.

[0020] The design of the state mapping algorithm converts the equipment coordinates and motion parameters of the physical production line into the pose data of the digital twin. Interpolation compensation technology is used to eliminate communication delays, ensuring consistency between virtual and real motion trajectories. A heartbeat mechanism is also established to periodically check the data consistency between the physical equipment and the virtual twin. When the deviation exceeds the threshold, an alarm is triggered to start the synchronization calibration process.

[0021] By embedding control logic into the digital twin, the state changes of the physical production line are analyzed in real time, and the analysis results of the virtual environment are fed back to the physical equipment through edge computing nodes, ensuring that the twin continuously reflects the behavior of the real production line.

[0022] A further improvement to the technical solution of this invention lies in the following: In step S3, the process of analyzing key performance indicators and simulating to verify feasibility and efficiency is as follows:

[0023] User-defined refactoring schemes (module additions and deletions, layout adjustments, etc.) are converted into machine-recognizable structured data, including module IDs, target locations, connection relationships, and process parameters. Through the parsing interface, these data are automatically mapped to the corresponding module instances in the production line digital twin model, ensuring that the refactoring instructions in the virtual environment strictly correspond to physical requirements. At the same time, data integrity is verified and conflicting configurations are eliminated.

[0024] The simulation of discrete events is initiated based on the digital twin engine. The module layout is dynamically adjusted according to the reconstruction scheme. Key performance indicators, including cycle time, collision probability and energy consumption, are calculated in real time. The feasibility of motion interference and dynamics is verified through the physics engine.

[0025] By comparing the simulation results with the preset key performance indicator thresholds, the feasibility and efficiency of the reconstruction scheme are evaluated. For conflicting or inefficient links, problem points are automatically marked and parameter adjustment suggestions are provided.

[0026] A further improvement to the technical solution of this invention lies in the following: the process of initiating discrete event simulation based on a digital twin engine is as follows:

[0027] Based on the topological relationship between the reconstructed modules, a virtual production line model framework is built in the digital twin engine. Through module ID mapping and coordinate transformation, the spatial layout initialization of each module is completed to ensure that the geometric position and connection relationship are consistent with the actual requirements. Process parameters, material properties and equipment motion constraints are loaded synchronously, and a discrete event simulation clock is configured and a time scaling factor is set to support rapid simulation and real-time debugging.

[0028] An event-driven mechanism is used to simulate the material flow path and equipment collaboration, dynamically triggering discrete events for processing, transmission, and detection;

[0029] The motion trajectory of the module is calculated in real time using a kinematic solver. Combined with a collision detection algorithm (bounding box detection), geometric interference risks are identified. Key performance indicators, including cycle time, collision probability, and energy consumption, are collected simultaneously. Sensor signal noise is eliminated through a digital filtering algorithm to ensure data reliability. During the simulation process, the equipment status and material position are dynamically updated through a visual interface, realizing full-dimensional monitoring and interactive debugging of the production process.

[0030] A further improvement to the technical solution of this invention lies in the following: In step S4, the process of iteratively updating the production line digital twin model until the feasibility and efficiency requirements are met is as follows:

[0031] Key performance indicators are extracted based on simulation results, and the links with excessive deviations are identified through statistical analysis. A rule engine is used to automatically locate the root cause of the problem, generate a structured diagnostic report, and clarify the specific parameters and modules that need to be optimized.

[0032] Based on the diagnostic results, the parameters of the reconstruction plan are revised, the geometric, motion and logical attributes of the production line digital twin model are updated synchronously, the virtual production line layout is quickly reconstructed through parametric modeling tools, and the physical engine constraints and discrete event rules are reconfigured to ensure that the model changes strictly match the actual needs.

[0033] Run the updated production line digital twin model for resimulation, compare the new results with the key performance indicators. If the results are not met, repeat the analysis-optimization process. If the results are met, freeze the model parameters and output the final reconstruction plan.

[0034] A further improvement to the technical solution of this invention lies in the following: the process of repeatable analysis-optimization and outputting the final reconstruction scheme is as follows:

[0035] Run the updated digital twin model of the production line, collect key performance indicators from the simulation output, including cycle time, collision probability and energy consumption, compare the simulation data with the preset key performance indicator thresholds, calculate the gap values ​​of each key performance indicator, and clarify the degree of deviation between the production line performance and the target.

[0036] If the simulation results do not reach the threshold of key performance indicators, analyze the causes of the gap, dig out the root causes of the problem from the perspective of production line process and resource allocation, formulate targeted optimization strategies based on the analysis results, adjust the model parameters and structure and simulate again. Through multiple iterations of analysis-optimization process, gradually narrow the gap with the threshold of key performance indicators, and update the diagnostic report after each iteration, reconfigure the simulation environment to ensure that the model behavior is consistent with the actual needs.

[0037] Repeated simulation verification and optimization adjustments are performed until the key performance indicators reach the preset key performance indicator thresholds three times consecutively. Once the target is met, the model parameters are frozen, and the final reconstruction scheme document is generated. This document includes the optimized geometric model, motion configuration, and complete simulation verification data chain. The performance improvement trajectory and parameter adjustment history of each iteration are recorded to form a traceable closed-loop optimization knowledge base.

[0038] A further improvement to the technical solution of the present invention is that S5 specifically includes:

[0039] Deploy a lightweight digital twin agent at the edge, allocate hardware resources and adapt communication protocols, configure module-level data acquisition rules, define the sampling frequency and transmission format of running data, establish a real-time communication link with the physical module, ensure low-latency data interaction, and synchronize initial model parameters to the digital twin.

[0040] The module's operational data is collected by the edge agent according to preset rules, the virtual model status is updated in real time in the digital twin, the threshold detection method is used for anomaly monitoring, a module-level status assessment report is generated, potential risk points are marked and an early warning mechanism is triggered.

[0041] Based on local monitoring results, the edge device calls the pre-set control strategy library to generate optimization instructions, and sends the instructions to the physical module for execution through a low-latency communication link. The digital twin status is updated synchronously to verify the control effect, forming a module-level closed-loop optimization cycle of acquisition-monitoring-execution.

[0042] A further improvement to the technical solution of this invention lies in the following: In step S6, the process of using AI algorithms to analyze and generate the optimal reconstruction strategy and then issuing it is as follows:

[0043] A global digital twin platform is built in the cloud, which aggregates real-time operation data from multiple production lines through standardized interfaces, covering data such as equipment operation and production processes. The data is cleaned, transformed, and standardized to eliminate noise and differences, and a global digital twin is established. It supports cross-production line topology mapping and status synchronization updates to ensure data consistency and traceability.

[0044] Based on the processed data, the pre-trained AI big model is called to calculate the reconstruction strategy. Combined with the constraints of equipment compatibility and capacity balance, the strategy is virtually verified through the global digital twin in the cloud. The verification results are fed back to the AI ​​big model to form a closed-loop optimization until the strategy meets the cross-production line collaboration indicators.

[0045] The generated optimal reconfiguration strategy is distributed to the control systems of each production line to ensure effective execution of the strategy. At the same time, data during the execution process is collected in real time and fed back to the cloud to evaluate the effectiveness of the strategy.

[0046] A further improvement to the technical solution of this invention lies in the following: In step S7, the process of completing the production line reconstruction and intelligent control is as follows:

[0047] The edge device receives the reconfiguration strategy from the cloud, parses it to generate device-level control instructions, and sends them to the PLC / robot controller in real time through a deterministic communication protocol. At the same time, the lightweight digital twin agent synchronously simulates the execution effect of the instructions, rehearses the triggering of physical device actions after no collision and no over-limit, and feeds back the execution status to the cloud.

[0048] Based on real-time data from edge devices, the cloud uses a large AI model to analyze deviations in the reconstruction strategy execution. The adjustment scheme is verified through digital twin simulation. The optimized reconstruction strategy is distributed in incremental configuration form to ensure low-bandwidth transmission efficiency, while updating the prediction model parameters of the global digital twin.

[0049] A two-way verification mechanism is formed between the edge and the cloud. The edge ensures the safe execution of instructions, while the cloud verifies the global effectiveness of the strategy. When the actual key performance indicators continue to deviate from the simulation prediction, the strategy rollback process is triggered until the system is in steady state, completing the partial or overall reconstruction and intelligent control of the production line.

[0050] Due to the adoption of the above technical solution, the technical progress achieved by this invention compared to the prior art is as follows:

[0051] 1. This invention provides a modular reconfigurable production line control system integration method. It constructs a virtual production line model through digital twin technology, supports the pre-playing and verification of reconfiguration schemes in a virtual environment, significantly reduces the trial and error time and cost required for traditional physical debugging, and further optimizes the reconfiguration strategy through cloud AI algorithms. Combined with real-time control at the edge, it realizes closed-loop optimization from virtual verification to physical execution, enabling the production line to quickly complete partial or overall reconfiguration according to production needs, and greatly improves the response speed and flexibility of the production line.

[0052] 2. This invention provides a modular and reconfigurable production line control system integration method. It collects physical production line data in real time through industrial protocol gateways and multi-source sensors. Combined with Kalman filtering and state mapping algorithms, it ensures the state synchronization accuracy between the digital twin and the physical equipment. Heartbeat detection and redundant communication mechanisms ensure the stability of the data link. In case of anomalies, it automatically triggers calibration or switching processes. The lightweight digital twin agent at the edge further enhances the module-level real-time monitoring capability and forms a two-way verification mechanism to effectively avoid control failures caused by communication delays or data deviations. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0054] Figure 1 This is a schematic diagram illustrating the workflow of the modular reconfigurable production line control system integration method of the present invention.

[0055] Figure 2 This is a schematic diagram of the method flow for the modular reconfigurable production line control system integration method of the present invention. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0057] Example 1, such as Figure 1 , Figure 2 As shown, this invention provides a modular reconfigurable production line control system integration method, comprising the following steps:

[0058] S1. Based on physical property parameters, establish a virtual mapping to form a basic digital twin model framework for the production line. Comprehensively collect physical property parameters of each module of the production line, covering dimensions, mass, range of motion, and dynamic characteristics. Systematically classify and organize the collected parameters, and use a unified data standard for structured storage to ensure data integrity and accuracy. Based on the organized physical property parameters, construct virtual models of each module according to the actual structure and module connection relationship of the production line using parametric modeling tools. Then assemble and integrate them to initially form a virtual model architecture for the production line. Verify the initially built virtual model architecture for the production line, check its matching degree with the actual production line in terms of structure and function, and make corrections and adjustments based on the verification results to form a basic digital twin model framework for the production line that reflects the characteristics of the production line.

[0059] The specific work involves: collecting physical attribute parameters for each module of the production line using 3D scanning, sensor measurements, and equipment manuals. This includes geometric dimensions (length, width, height, interface location), mass distribution (weight, center of gravity), range of motion (displacement, rotation angle limits), and dynamic characteristics (motor power, torque curve, acceleration). After collection, the physical attribute parameters are systematically classified according to module type to ensure data clarity. A unified data standard is used to structure and store the classified parameters, establishing a parameter database to guarantee data integrity and accuracy, and avoid data loss or errors. Based on the compiled physical attribute parameters, parametric modeling tools (CAD) are used to begin constructing a virtual model of the production line. Virtual models of each module are built according to the actual structure of the production line and the module connection relationships. The size and shape parameters of each module are defined to ensure consistency with the actual module. After the modeling of a single module is completed, the modules are assembled and integrated according to the production line layout. The connection methods and relative positional relationships between the modules are clarified, and a preliminary virtual model architecture for the production line is built. The preliminary virtual model architecture is fully verified to check its matching degree with the actual production line in terms of structure and function. Structurally, the module layout and connection methods of the virtual model are compared with those of the actual production line to see if they are consistent. From a functional perspective, the operation of the production line is simulated to observe whether the movement state and collaborative work of each module in the virtual model are consistent with reality. Based on the verification results, the deviations and problems in the virtual model are corrected and adjusted, the module parameters are optimized, and the connection relationships are adjusted. After repeated verification and correction, a basic digital twin model framework for the production line that can accurately reflect the characteristics of the production line is formed.

[0060] S2. Based on real-time data acquisition from field sensors, establish a dynamic mapping mechanism for a digital twin synchronized with the physical production line status. Deploy an industrial protocol gateway on the physical production line to connect various field sensors to PLCs and robot controllers, establish data acquisition channels, and configure a time-series database to store status data output by field sensors, including position, speed, and temperature. Define a data tagging system to give each parameter a unique identifier. Design a status mapping algorithm to convert the equipment coordinates and motion parameters of the physical production line into pose data of the digital twin. Use interpolation compensation technology to eliminate communication delays and ensure consistency between virtual and real motion trajectories. Establish a heartbeat mechanism to periodically verify the data consistency between the physical equipment and the virtual twin. When the deviation exceeds a threshold, trigger an alarm to start the synchronization calibration process. Embed control logic in the digital twin to analyze the status changes of the physical production line in real time, and feed back the analysis results of the virtual environment to the physical equipment through edge computing nodes to ensure that the twin continuously reflects the behavior of the real production line.

[0061] The specific work involves: deploying industrial protocol gateways on the physical production line, compatible with multiple communication protocols, to achieve data integration from PLCs, robot controllers, and field sensors. The industrial protocol gateways collect real-time status data such as position, speed, and temperature through a unified interface, and identify parameters according to a predefined tagging system to ensure the uniqueness and traceability of the data source. The collected data is transmitted to a time-series database via a high-throughput communication link, supporting millisecond-level writing and compressed storage to meet the long-term archiving requirements of high-frequency data. Simultaneously, a data quality control mechanism is established to filter and repair outliers and missing values ​​in real time; and a state mapping algorithm based on coordinate transformation and kinematic models is designed to map the physical... The coordinates and motion parameters of the equipment on the production line are converted into virtual representations of the digital twin. Kalman filtering technology is used to compensate for data deviations, ensuring the synchronization of virtual and real motion trajectories. A heartbeat detection mechanism is introduced to periodically compare key state parameters of the physical equipment and the digital twin. If the deviation exceeds a preset threshold, a synchronization calibration process is triggered, forcibly resetting the virtual model state or adjusting the control parameters of the physical equipment to maintain the consistency of the virtual and real systems. A control logic module is embedded in the digital twin to analyze the state changes of the physical production line in real time. The analysis results of the virtual environment are fed back to the PLC or robot controller through edge computing nodes, forming a two-way control closed loop to ensure the stable operation of the system during dynamic reconstruction.

[0062] S3. Import the reconfiguration scheme into the virtual environment, simulate the module combination process through the production line digital twin model, analyze key performance indicators, and verify the feasibility and efficiency through simulation. Convert the user-defined reconfiguration scheme (module addition, deletion, layout adjustment, etc.) into machine-recognizable structured data, including module ID, target location, connection relationship and process parameters. Automatically map to the corresponding module instance in the production line digital twin model through the parsing interface to ensure that the reconfiguration instructions in the virtual environment strictly correspond to the physical requirements. At the same time, verify the data integrity, eliminate conflicting configurations, start discrete event simulation based on the digital twin engine, dynamically adjust the module layout according to the reconfiguration scheme, calculate key performance indicators including cycle time, collision probability and energy consumption in real time, and verify the feasibility of motion interference and dynamics through the physics engine. Compare the simulation results with the preset key performance indicator thresholds to evaluate the feasibility and efficiency of the reconfiguration scheme. For conflicting or inefficient links, automatically mark the problem points and provide parameter adjustment suggestions.

[0063] Furthermore, the process of initiating discrete event simulation based on the digital twin engine is as follows:

[0064] Based on the topological relationships between the reconstructed modules, a virtual production line model framework is constructed in the digital twin engine. Through module ID mapping and coordinate transformation, the spatial layout initialization of each module is completed, ensuring that the geometric position, connection relationship and actual needs are consistent. Process parameters, material properties and equipment motion constraints are loaded synchronously, and a discrete event simulation clock is configured and a time scaling factor is set to support rapid simulation and real-time debugging. The physics engine calculates the dynamic characteristics of the modules in real time and integrates a multi-physics coupling model to ensure the consistency of the behavior of the virtual environment and the physical production line. An event-driven mechanism is used to simulate the material flow path and equipment collaborative operation, dynamically triggering discrete events of processing, transmission and detection. The kinematic solver is used to calculate the motion trajectory of the modules in real time, and a collision detection algorithm (bounding box detection) is used to identify geometric interference risks. Key performance indicators including cycle time, collision probability and energy consumption are collected synchronously. Sensor signal noise is eliminated through digital filtering algorithm to ensure data reliability. During the simulation process, the equipment status and material position are dynamically updated through a visual interface to realize full-dimensional monitoring and interactive debugging of the production process.

[0065] The specific tasks include: converting user-defined refactoring schemes into machine-readable structured data; defining a standardized data model covering key fields such as module ID, target coordinates, connection relationships, and process parameters; mapping user input to module instances in the production line digital twin model through a parsing interface to ensure strict matching between adjustment instructions in the virtual environment and physical requirements; verifying data integrity, such as checking for missing required parameters and valid module IDs, and eliminating conflicting configurations; based on the refactoring scheme, the digital twin engine initiates discrete event simulation, dynamically adjusts the module layout, simulates production line operation, calculates module motion trajectories through a physics engine, detects potential geometric interference and dynamic conflicts, and synchronously simulates material flow and equipment collaborative operation in real time. The simulation process is designed to measure key performance indicators such as cycle time, equipment utilization, and energy consumption. It must support time scaling to allow for both rapid global evaluation and pause for detailed inspection. Specifically, it utilizes collision detection algorithms to identify spatial conflicts and employs kinematic solvers to verify joint limits and acceleration constraints. The simulation results are compared with preset key performance indicator thresholds to quantify the feasibility of the reconstruction scheme. Inefficient components or conflict points are automatically identified, and specific parameter deviations are marked. Optimization suggestions are generated based on a rule engine, including adjusting module spacing, optimizing motion trajectories, or redistributing process loads. For infeasible solutions, root cause analysis is provided. All analysis results are stored in a structured format. The final output includes pass / fail criteria, a key performance indicator comparison table, and a list of recommended optimization parameters.

[0066] S4. Optimize the reconstruction plan based on the simulation results, iteratively update the production line digital twin model until the feasibility and efficiency requirements are met, extract key performance indicators based on the simulation results, identify the links with excessive deviations through statistical analysis, use a rule engine to automatically locate the root cause of the problem, generate a structured diagnostic report, clarify the specific parameters and modules that need to be optimized, correct the parameters of the reconstruction plan based on the diagnostic results, and update the geometric, motion and logical attributes of the production line digital twin model simultaneously. Quickly reconstruct the virtual production line layout through parametric modeling tools, reconfigure the physical engine constraints and discrete event rules to ensure that the model changes strictly match the actual needs, run the updated production line digital twin model for re-simulation, compare the gap between the new results and the key performance indicators, if the standards are not met, repeat the analysis-optimization process, if the standards are met, freeze the model parameters and output the final reconstruction plan;

[0067] Furthermore, the process of repeating the analysis-optimization workflow and outputting the final refactoring solution is as follows:

[0068] Run the updated digital twin model of the production line, collect key performance indicators (KPIs) from the simulation output, including cycle time, collision probability, and energy consumption, and compare the simulation data with preset KPI thresholds. Calculate the gap values ​​for each KPI to clarify the degree of deviation between the production line performance and the target. If the simulation results do not reach the KPI thresholds, analyze the causes of the gap, dig out the root causes from the perspective of production line processes and resource allocation, formulate targeted optimization strategies based on the analysis results, adjust the model parameters and structure, and simulate again. Through multiple iterations of analysis and optimization processes, gradually narrow the gap with the KPI thresholds. Update the diagnostic report after each iteration, reconfigure the simulation environment to ensure that the model behavior is consistent with actual needs, and repeat simulation verification and optimization adjustments until the KPIs reach the preset KPI thresholds three times consecutively. The solution is then deemed satisfactory. Once satisfactory, the model parameters are frozen, and the final reconstructed solution document is generated, including the optimized geometric model, motion configuration, and complete simulation verification data chain. Record the performance improvement trajectory and parameter adjustment history of each iteration to form a traceable closed-loop optimization knowledge base.

[0069] The specific work involves: extracting key performance indicators (KPIs) based on multi-dimensional data from the production line digital twin simulation using statistical process control methods, including cycle time, collision probability, and energy consumption; identifying KPI deviations exceeding limits using control chart analysis; quantifying the significance of performance differences between modules using variance analysis; automatically locating the root cause of problems to specific modules and parameters by matching a predefined problem pattern library with a rule engine; generating a structured diagnostic report that clarifies the types of parameters requiring optimization and related modules; and revising the reconstruction plan based on the optimization direction determined in the diagnostic report using parametric modeling tools, covering geometric attributes (module dimensions, connection point coordinates), motion characteristics (velocity / acceleration curves, joint limits), and logical rules (event triggering conditions, priority allocation), while simultaneously updating the production line digital twin model. Feature tree technology is used to achieve rapid propagation of geometric changes. Kinematic remapping ensures dynamic adaptation of constraints. The contact parameters (friction coefficient, elastic modulus) and discrete event simulation rules (material delivery interval, equipment start-up and shutdown logic) of the physics engine are reconfigured to ensure that the model behavior is strictly consistent with actual requirements. The updated production line digital twin model is run for multiple rounds of simulation, continuously collecting data on new key performance indicators and comparing them with preset key performance indicator thresholds. If the target is not met, a closed-loop optimization process is initiated to optimize parameters and adjust the module layout synchronously. The diagnostic report is updated after each iteration. When the simulation results meet all key performance indicator threshold requirements for three consecutive times, the solution is deemed to be up to standard. The model parameters are frozen and the final reconstruction solution document is generated, which includes the optimized geometric model, motion configuration, and simulation verification data chain.

[0070] S5. Deploy a lightweight digital twin agent at the edge to perform module-level real-time data collection, status monitoring, and localized control command issuance;

[0071] S6. Build a global digital twin platform in the cloud, integrate data from multiple production lines, use AI algorithms to analyze and generate the optimal reconstruction strategy and distribute it.

[0072] S7: Through real-time interaction between the edge and cloud, control commands are executed at the edge, and strategies are dynamically adjusted in the cloud to complete production line reconstruction and intelligent control.

[0073] Example 2, as Figure 1 , Figure 2 As shown, based on Embodiment 1, the present invention provides a technical solution: Preferably, S5 specifically includes:

[0074] Deploy a lightweight digital twin agent at the edge to allocate hardware resources and adapt communication protocols, configure module-level data acquisition rules, define the sampling frequency and transmission format of running data, establish a real-time communication link with the physical module to ensure low-latency data interaction, synchronize initial model parameters to the digital twin, collect module running data through the edge agent according to preset rules, update the virtual model status in the digital twin in real time, use threshold detection method for anomaly monitoring, generate module-level status assessment reports, mark potential risk points and trigger early warning mechanisms, and based on local monitoring results, call the preset control strategy library at the edge to generate optimization instructions, send the instructions to the physical module for execution through the low-latency communication link, and synchronously update the digital twin status to verify the control effect, forming a module-level closed-loop optimization cycle of acquisition-monitoring-execution.

[0075] The specific tasks include: deploying a lightweight digital twin agent at the edge, allocating hardware resources including CPU core binding, memory reservation, and real-time task priority settings to ensure computing resources meet low-latency requirements, adapting to various industrial communication protocols, configuring data acquisition rules, defining the sampling frequency and transmission format of runtime data, establishing a real-time communication link with the physical module, using time-sensitive networking technology to ensure data interaction latency ≤10ms, synchronizing initial model parameters to the digital twin to ensure consistency between the virtual model and the initial state of the physical entity, deploying a heartbeat detection mechanism to periodically verify the stability of the communication link, and automatically switching to a redundant channel in case of anomalies; the edge agent collects module-level runtime data according to preset rules, and transmits the data through a lightweight state... The update algorithm maps physical state changes in real time in the digital twin, uses a multi-level threshold detection method to monitor key parameters, marks potential risk points based on a rule engine, triggers a graded early warning mechanism, including log alarms, audible and visual alarms, and emergency shutdown, and stores the complete anomaly context through a time-series database to support post-event root cause analysis; the edge has a built-in lightweight control strategy library, matches the optimal strategy based on real-time monitoring results, generates optimization instructions, and sends the optimization instructions to the physical module through a deterministic communication protocol. The digital twin synchronously simulates the effect of the instructions, pre-plays the feasibility of the actions through the physical engine, and confirms the effectiveness of the instructions after verifying no collisions and no exceeding limits. The execution results are fed back to the digital twin for state calibration, forming a closed-loop optimization cycle of acquisition-monitoring-execution;

[0076] In S6, the process of using AI algorithms to analyze, generate, and distribute the optimal reconstruction strategy is as follows:

[0077] A global digital twin platform is built in the cloud, aggregating real-time operational data from multiple production lines through standardized interfaces. This data covers equipment operation, production processes, and other aspects. The data is cleaned, transformed, and standardized to eliminate noise and discrepancies, creating a global digital twin that supports cross-production line topology mapping and state synchronization updates. This ensures data consistency and traceability. Based on the processed data, a pre-trained AI model is invoked to calculate a reconstruction strategy. This strategy is combined with constraints such as equipment compatibility and capacity balance, and then virtually verified through the global digital twin in the cloud. The verification results are fed back to the AI ​​model to form a closed-loop optimization until the strategy meets cross-production line collaboration indicators. The generated optimal reconstruction strategy is then distributed to the control systems of each production line to ensure effective execution. Simultaneously, data during the execution process is collected in real time and fed back to the cloud to evaluate the strategy's effectiveness.

[0078] The specific work involves: building a global digital twin platform in the cloud; aggregating real-time operational data from multiple production lines through standardized RESTful / gRPC interfaces, including heterogeneous data sources such as equipment status and production processes; processing the raw data using a data cleaning pipeline, performing noise reduction, format conversion, and unit standardization to eliminate data deviations caused by differences in equipment models or acquisition protocols; constructing a global digital twin based on the cleaned data; achieving cross-production line equipment association mapping through a topology graph; and using an optimistic locking mechanism to ensure synchronized status updates in a distributed environment. Based on the standardized data, a pre-trained AI model is invoked, with inputs including equipment compatibility matrices and capacity balance constraints. After the model outputs a preliminary reconstruction strategy, virtual verification is performed through the global digital twin in the cloud. Discrete event simulation verifies logistics efficiency, and a physics engine detects operational data. The system dynamically intervenes, records verification metrics in a time-series database, and feeds back the verification results to the AI ​​big model in the form of a loss function. Iterative optimization of strategy parameters continues until cross-production line collaboration metrics are met. Each iteration generates a strategy feasibility report, recording constraint violations and optimization trajectories. The optimal reconstruction strategy generated in the cloud is distributed to the edge controllers of each production line through an encrypted channel. The optimal reconstruction strategy includes a module layout diagram, motion parameter configuration file, and control logic update package. Upon receiving the data, the edge controller immediately verifies the digital signature and performs pre-execution verification for compatibility in a secure sandbox. After the optimal reconstruction strategy officially takes effect, the global digital twin platform collects execution data in real time. The strategy effectiveness is evaluated by comparing the deviation rate between actual key performance indicators and simulation prediction values. When the deviation exceeds the threshold, the strategy is rolled back. Simultaneously, the feedback data is injected into the training set of the AI ​​big model to continuously optimize the model's generalization ability.

[0079] In S7, the process of completing production line restructuring and intelligent control is as follows:

[0080] The edge device receives the reconfiguration strategy from the cloud, parses it to generate device-level control instructions, and sends them to the PLC / robot controller in real time via a deterministic communication protocol. At the same time, the lightweight digital twin agent synchronously simulates the execution effect of the instructions, rehearses the triggering of physical equipment actions after collision-free and limit-free operation, and feeds back the execution status to the cloud. Based on the real-time data fed back from the edge device, the cloud calls the AI ​​large model to analyze the execution deviation of the reconfiguration strategy, verifies the adjustment scheme through digital twin simulation, and sends the optimized reconfiguration strategy in the form of incremental configuration to ensure low-bandwidth transmission efficiency. At the same time, it updates the prediction model parameters of the global digital twin. The edge device and the cloud form a two-way verification mechanism, that is, the edge device ensures the safe execution of instructions, and the cloud verifies the global effectiveness of the strategy. When the actual key performance indicators continue to deviate from the simulation prediction, the strategy rollback process is triggered until the system is in steady state, completing the partial or overall reconfiguration and intelligent control of the production line.

[0081] The specific tasks are as follows: After receiving the reconstructing strategy transmitted encrypted from the cloud, the edge device verifies the data integrity using digital signatures and parses it to generate device-level control commands. A deterministic communication protocol is used to send the device-level control commands to the PLC / robot controller in real time, ensuring timing determinism. Simultaneously, a lightweight digital twin agent at the edge device synchronously simulates the execution effect of the device-level control commands. A physics engine detects risks such as motion interference and overload. Physical device actions are triggered only when virtual pre-simulation confirms no collisions and parameters are within safety thresholds. During execution, the edge device continuously collects real-time information on device status and sensor data, and feeds it back to the cloud using a high-efficiency compression algorithm. Based on the real-time data from the edge device, the cloud uses an AI model to analyze the execution deviation of the reconstructing strategy, identifies the root causes of differences between key performance indicators and simulation predictions, verifies the adjustment scheme through digital twin simulation, simulates the impact of different parameter combinations on global efficiency, and generates an optimized incremental configuration package for the reconstructing strategy. Only the changed parts are transmitted to reduce bandwidth usage. Strategy parameters are synchronously updated to the global digital twin to improve its prediction accuracy. If the deviation still exceeds the threshold after multiple iterations, the strategy is marked as pending rollback and a deep analysis process is triggered. A two-way verification mechanism is built between the edge and the cloud: the edge is responsible for instruction-level security, ensuring risk-free execution of a single device through lightweight twin pre-simulation and hardware security modules; the cloud is responsible for strategy-level global effectiveness verification, analyzing cross-device collaboration effects based on multi-production line data aggregation. When actual key performance indicators continuously deviate from simulation predictions, the cloud automatically triggers the strategy rollback process, restoring to the previous stable version and generating a deviation root cause report. The system continuously optimizes by dynamically adjusting the weights of the AI ​​large model and updating constraint parameters until the production line operation converges to a steady state. When the steady state is finally reached, the control parameters, topology relationships, and performance data of all modules are automatically archived to form a reusable reconstructed knowledge graph.

[0082] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A modular reconfigurable production line control system integration method, characterized in that, Includes the following steps: S1. Establish a virtual mapping based on physical property parameters to form a basic model framework for the production line digital twin; S2. Based on real-time data acquisition from on-site sensors, establish a dynamic mapping mechanism for a digital twin that is synchronized with the physical production line status; S3. Import the reconstruction scheme into the virtual environment, simulate the module combination process through the production line digital twin model, analyze key performance indicators, and verify the feasibility and efficiency through simulation. S4. Optimize the reconstruction plan based on the simulation results, and iteratively update the production line digital twin model until the feasibility and efficiency requirements are met. S5. Deploy a lightweight digital twin agent at the edge to perform module-level real-time data collection, status monitoring, and localized control command issuance; S6. Build a global digital twin platform in the cloud, integrate data from multiple production lines, use AI algorithms to analyze and generate the optimal reconstruction strategy and distribute it. S7. Through real-time interaction between the edge and cloud, control commands are executed at the edge, and strategies are dynamically adjusted in the cloud to complete production line reconstruction and intelligent control. In S1, the formation process of the basic model framework for the production line digital twin is as follows: Comprehensive collection of physical property parameters of each module of the production line, covering size, mass, range of motion and dynamic characteristics; systematic classification and organization of the collected parameters; and structured storage using a unified data standard. Based on the organized physical property parameters, and using parametric modeling tools, virtual models of each module are constructed according to the actual structure and module connection relationships of the production line. These models are then assembled and integrated to form a preliminary virtual model architecture for the production line. The initial virtual model architecture of the production line is verified to check its matching degree with the actual production line in terms of structure and function. Based on the verification results, it is corrected and adjusted to form a basic digital twin model framework of the production line that reflects the characteristics of the production line. In step S4, the process of iteratively updating the production line digital twin model until the feasibility and efficiency requirements are met is as follows: Key performance indicators are extracted based on simulation results, and the links with excessive deviations are identified through statistical analysis. A rule engine is used to automatically locate the root cause of the problem, generate a structured diagnostic report, and clarify the specific parameters and modules that need to be optimized. Based on the diagnostic results, the parameters of the reconstruction plan are corrected, the geometric, motion and logical attributes of the production line digital twin model are updated synchronously, the virtual production line layout is quickly reconstructed through parametric modeling tools, and the physical engine constraints and discrete event rules are reconfigured. Run the updated production line digital twin model for resimulation, compare the new results with the key performance indicators. If the results are not met, repeat the analysis-optimization process. If the results are met, freeze the model parameters and output the final reconstruction plan.

2. The modular reconfigurable production line control system integration method according to claim 1, characterized in that: In S2, the process of establishing a dynamic mapping mechanism for a digital twin synchronized with the physical production line status is as follows: Deploy industrial protocol gateways on the physical production line to connect various field sensors to PLCs and robot controllers, establish data acquisition channels, configure a time series database to store status data output by field sensors, including position, speed, and temperature, and define a data tagging system to give each parameter a unique identifier. The design of the state mapping algorithm converts the equipment coordinates and motion parameters of the physical production line into the pose data of the digital twin. Interpolation compensation technology is used to eliminate communication delays, and a heartbeat mechanism is established to periodically check the data consistency between the physical equipment and the virtual twin. When the deviation exceeds the threshold, an alarm is triggered to start the synchronization calibration process. Control logic is embedded in the digital twin to analyze the status changes of the physical production line in real time, and the analysis results of the virtual environment are fed back to the physical equipment through edge computing nodes.

3. The modular reconfigurable production line control system integration method according to claim 1, characterized in that: In S3, the process of analyzing key performance indicators and simulating to verify feasibility and efficiency is as follows: The user-defined refactoring scheme is converted into machine-recognizable structured data, including module ID, target location, connection relationship and process parameters. It is automatically mapped to the corresponding module instance of the production line digital twin model through the parsing interface, while verifying data integrity and eliminating conflicting configurations. The simulation of discrete events is initiated based on the digital twin engine. The module layout is dynamically adjusted according to the reconstruction scheme. Key performance indicators, including cycle time, collision probability and energy consumption, are calculated in real time. The feasibility of motion interference and dynamics is verified through the physics engine. By comparing the simulation results with the preset key performance indicator thresholds, the feasibility and efficiency of the reconstruction scheme are evaluated. For conflicting or inefficient links, problem points are automatically marked and parameter adjustment suggestions are provided.

4. The modular reconfigurable production line control system integration method according to claim 3, characterized in that: The process of initiating discrete event simulation based on the digital twin engine is as follows: Based on the topological relationship between the reconstructed modules, a virtual production line model framework is built in the digital twin engine. Through module ID mapping and coordinate transformation, the spatial layout initialization of each module is completed, process parameters, material properties and equipment motion constraints are loaded synchronously, and a discrete event simulation clock is configured and a time scaling factor is set. An event-driven mechanism is used to simulate the material flow path and equipment collaboration, dynamically triggering discrete events for processing, transmission, and detection; The motion trajectory of the module is calculated in real time using a kinematic solver, and the geometric interference risk is identified by a collision detection algorithm. Key performance indicators, including cycle time, collision probability and energy consumption, are collected simultaneously. Sensor signal noise is eliminated by a digital filtering algorithm. During the simulation process, the equipment status and material position are dynamically updated through a visual interface, realizing full-dimensional monitoring and interactive debugging of the production process.

5. The modular reconfigurable production line control system integration method according to claim 1, characterized in that: The process of repeatability analysis-optimization workflow and outputting the final reconstruction solution is as follows: Run the updated digital twin model of the production line, collect key performance indicators from the simulation output, including cycle time, collision probability and energy consumption, compare the simulation data with the preset key performance indicator thresholds, calculate the gap values ​​of each key performance indicator, and clarify the degree of deviation between the production line performance and the target. If the simulation results do not reach the key performance indicator threshold, analyze the causes of the gap, formulate targeted optimization strategies based on the analysis results, adjust the model parameters and structure, and simulate again. Through multiple iterations of analysis-optimization process, gradually narrow the gap with the key performance indicator threshold, and update the diagnostic report and reconfigure the simulation environment after each iteration. Repeated simulation verification and optimization adjustments are performed until the key performance indicators reach the preset key performance indicator thresholds three times consecutively. Once the target is met, the model parameters are frozen, and the final reconstruction scheme document is generated. This document includes the optimized geometric model, motion configuration, and complete simulation verification data chain. The performance improvement trajectory and parameter adjustment history of each iteration are recorded to form a traceable closed-loop optimization knowledge base.

6. The modular reconfigurable production line control system integration method according to claim 5, characterized in that: S5 specifically includes: Deploy a lightweight digital twin agent at the edge, allocate hardware resources and adapt communication protocols, configure module-level data acquisition rules, define the sampling frequency and transmission format of running data, establish a real-time communication link with the physical module, and synchronize the initial model parameters to the digital twin. The module's operational data is collected by the edge agent according to preset rules, the virtual model status is updated in real time in the digital twin, the threshold detection method is used for anomaly monitoring, a module-level status assessment report is generated, potential risk points are marked and an early warning mechanism is triggered. Based on local monitoring results, the edge device calls the pre-set control strategy library to generate optimization instructions, and sends the instructions to the physical module for execution through a low-latency communication link. The digital twin status is updated synchronously to verify the control effect, forming a module-level closed-loop optimization cycle of acquisition-monitoring-execution.

7. The modular reconfigurable production line control system integration method according to claim 6, characterized in that: In step S6, the process of using AI algorithms to analyze and generate the optimal reconstruction strategy and then issuing it is as follows: A global digital twin platform is built in the cloud, which aggregates real-time operation data from multiple production lines through standardized interfaces, cleans, transforms and standardizes the data, and establishes a global digital twin. Based on the processed data, the pre-trained AI big model is called to calculate the reconstruction strategy. Combined with the constraints of equipment compatibility and capacity balance, the strategy is virtually verified through the global digital twin in the cloud. The verification results are fed back to the AI ​​big model to form a closed-loop optimization until the strategy meets the cross-production line collaboration indicators. The generated optimal reconfiguration strategy is distributed to the control systems of each production line. At the same time, data during the execution process is collected in real time and fed back to the cloud to evaluate the effectiveness of the strategy.

8. The modular reconfigurable production line control system integration method according to claim 7, characterized in that: In S7, the process of completing production line restructuring and intelligent control is as follows: The edge device receives the reconfiguration strategy from the cloud, parses it to generate device-level control instructions, and sends them to the PLC / robot controller in real time through a deterministic communication protocol. At the same time, the lightweight digital twin agent synchronously simulates the execution effect of the instructions, rehearses the triggering of physical device actions after no collision and no over-limit, and feeds back the execution status to the cloud. Based on real-time data from edge devices, the cloud uses a large AI model to analyze deviations in the reconstruction strategy execution. The adjustment scheme is verified through digital twin simulation. The optimized reconstruction strategy is distributed in the form of incremental configuration, while the prediction model parameters of the global digital twin are updated. A two-way verification mechanism is formed between the edge and the cloud. When the actual key performance indicators continue to deviate from the simulation prediction, the strategy rollback process is triggered until the system is in steady state, completing the partial or overall reconstruction and intelligent control of the production line.

Images