Flexible processing production line-oriented digital twin platform bidirectional synchronization method and system

Abstract

The invention discloses a flexible processing production line-oriented digital twin platform bidirectional synchronization method and system, and the method comprises the steps: converting static process data into a dynamic event flow capable of driving simulation through a CAM semantic adapter, and breaking through and activating a'process-simulation-twin 'full link; incremental synchronization based on simulation logic time and client interpolation rendering are adopted, millimeter-level synchronization of visual performance and a logic process can still be ensured in a complex scene, and synchronization and visualization with high fidelity and strict time sequence are achieved. Actual time, takt and concurrency relations are given to a static process in a simulation platform, and engineering analysis capability of digital twinning depth is given, so that quantitative analysis of production takt and productivity bottleneck becomes possible; through an abstract feedback instruction and a routing mechanism, an extensible virtuous cycle of simulation verification-visual analysis-decision optimization is formed, and meanwhile, the stability of a core time sequence link and the overall resource efficiency of the system are guaranteed.

Description

Technical Field

[0001] This invention relates to the field of industrial digital twin and intelligent manufacturing technology, and is particularly applicable to a two-way synchronization method and system for digital twin platforms for flexible manufacturing production lines. Background Technology

[0002] Flexible manufacturing lines, as core production carriers in discrete manufacturing, aerospace, and other fields, directly determine a company's market competitiveness through their level of intelligence. They are a key support for the manufacturing industry's transformation from traditional rigid production models to flexible manufacturing characterized by "small batches, multiple varieties, and rapid response." Digital twin technology, by constructing a closed-loop system of "physical production line - virtual model - twin data - application services," provides an integrated technical solution for the full lifecycle management of flexible production lines and has become a core enabling technology driving the digital transformation of the manufacturing industry.

[0003] Currently, digital twin technology has been widely applied and researched in various fields such as intelligent workshops, power systems, and smart cities. However, significant technological bottlenecks remain in the multi-software integration scenarios of flexible manufacturing production lines. Building a high-fidelity, interactive, and decision-making-capable digital twin of a flexible manufacturing production line often requires the integration of multiple software tools, including CAM systems, discrete event simulation platforms, and 3D visualization platforms. Ideally, it should achieve end-to-end data connectivity and intelligent closed-loop integration across the entire process of "process-simulation-twin." However, current technologies face severe challenges in practical integration, hindering the full realization of the value of digital twins, specifically as follows: 1. Data Semantic Heterogeneity and Rigid Conversion: The NC / G code output by the CAM system focuses on toolpath and machining logic, the simulation platform uses an object-oriented event-driven model and logistics data, while the twin platform relies on 3D graphics rendering data streams. The data structures and business semantics of these three systems are drastically different. Existing integration solutions mostly use customized "point-to-point" data conversion scripts or intermediate files. This approach is highly dependent on the specific workpiece process; whenever the processing object on the production line changes (e.g., from a turbine disk to a casing), the conversion logic must be redeveloped or adjusted, lacking versatility and flexibility, and the conversion process is prone to introducing errors.

[0004] 2. The synchronization mechanism is crude and unidirectional: Mainstream methods achieve synchronization through periodic polling or simple state triggers, which cannot accurately respond to discrete events with clear business implications, such as "process completion," "tool life exhaustion," and "AGV precise approach." This results in asynchrony or delay between the virtual world and the physical world (or high-precision simulation logic) at key business nodes, reducing the reliability and credibility of the digital twin model. More importantly, the data flow is usually a unidirectional transfer from CAM to simulation and then to visualization, forming an "open loop." When the visualization twin discovers potential problems or optimization points through its built-in intelligent analysis models (such as machine learning-based fault prediction and bottleneck identification), there is a lack of standardized and automated ways to feed the decision information back to the upstream simulation platform for strategy adjustment or to guide the optimization of process parameters in the CAM system. This prevents the formation of a closed loop of "perception-analysis-decision-execution" in the digital twin.

[0005] 3. Low data transmission efficiency: To maintain the real-time state of the virtual model, it is often necessary to transmit massive amounts of data such as device coordinates and speed at high frequency, which contains a large amount of redundant information (such as repeated coordinates when the device is stationary). Using standard industrial communication protocols (such as OPC UA) for full and periodic broadcasting will result in high network bandwidth utilization and large transmission latency, making it difficult to meet the requirements of applications with stringent immediacy requirements such as virtual debugging and real-time remote monitoring.

[0006] Therefore, for flexible manufacturing production lines, a specific scenario with extremely high requirements for dynamic adjustment and rapid response, there is an urgent need for an innovative integration method to solve the semantic interoperability between heterogeneous platforms and to build an efficient, high-fidelity data synchronization pipeline that supports two-way interaction, thereby unleashing the full potential of digital twins for flexible production lines. Summary of the Invention

[0007] The purpose of this invention is to provide a two-way synchronization method and system for a digital twin platform for flexible manufacturing lines, which solves the problems of current digital twin systems being unable to accurately respond to key business nodes, having unidirectional data flow, and being unable to form a closed loop of perception-analysis-decision-execution.

[0008] To achieve the above objectives, the core concept of the bidirectional synchronization method and system for digital twin platforms for flexible manufacturing lines described in this invention is to use an existing general data service platform as the core hub, extract static process blueprints by extending the CAM semantic adapter, assign time-series dimensions through the simulation platform, form a data pipeline of process planning-dynamic simulation-intelligent twin, and realize an enhanced integrated framework of high-fidelity dynamic data, synchronization and bidirectional feedback of abstract events through enhanced twin synchronization services.

[0009] The general-purpose data service platform is a centralized data hub integrating multi-protocol adaptation (such as OPC UA, MQTT), real-time message bus (such as WebSocket), and data storage (such as MySQL). In intelligent manufacturing production lines, the platform connects CNC machine tools, industrial robots, PLCs, sensors, AGVs, and various new and old equipment. Through multi-protocol adaptation, it aggregates data from intelligent manufacturing equipment in real time, utilizes the real-time message bus for high-speed flow and distribution, and categorizes and stores the data in appropriate databases. This enables transparent monitoring of the entire production process and provides a data foundation for quality traceability and process optimization. Simultaneously, through standard APIs, it provides unified and reliable data services to various business applications, such as MES and digital twins, driving business decisions and innovation, and improving production line efficiency and flexibility.

[0010] The bidirectional synchronization method for a digital twin platform for flexible manufacturing lines described in this invention, based on a general data service platform, includes the following steps: S1, develop and deploy CAM semantic adapters to semantically parse CAM data and encapsulate it into standardized process events before injecting it into a general data service platform; S2, the simulation platform simulates the packaged standardized process events, generates a time-series data stream, and injects it into the general data service platform; S3 is used to develop and deploy enhanced twin synchronization services, subscribe to and receive time-series data streams in real time, compare new time-series data with the previous state snapshot of the corresponding entity, and generate incremental update messages. S4, the digital twin client drives scene updates based on incremental update messages, outputs standardized feedback instructions through user interaction or intelligent analysis, and sends them back to the general data service platform through a two-way communication link; S5, the general data service platform, distributes feedback instructions to downstream systems based on the type and target of the instructions.

[0011] Furthermore, the time-series data stream consists of state changes, logical events, and motion steps with simulation timestamps, generated by the simulation platform driving the virtual production line model based on standardized process events.

[0012] Furthermore, the incremental update message is the state difference value between the current time-series data stream and the previous state snapshot of the corresponding entity.

[0013] Furthermore, an intelligent redundancy filtering strategy is adopted, which only releases data to the general data service platform when the data changes exceed a preset threshold.

[0014] Furthermore, the feedback instruction types include control instructions and optimization suggestions for process root causes; the control instructions are distributed to the simulation platform; and the optimization suggestions for process root causes are distributed to the CAM system.

[0015] The present invention discloses a bidirectional synchronization system for a digital twin platform for flexible manufacturing production lines, based on a general data service platform, comprising: The CAM semantic adapter module is used to parse and map the machining data of the CAM system into standardized process semantic units and encapsulate them into standardized process events. The timing-driven module of the simulation platform drives the dynamic execution of standardized process events within the simulation platform, generating a timing-driven data stream. The enhanced twin synchronization service module compares the time-series data stream with the previous state snapshot of the corresponding entity attribute by attribute, generates incremental update messages, and synchronizes them to the digital twin platform. The intelligent twin client module renders the scene based on incremental update messages, and converts optimization suggestions generated by user interaction or local intelligent analysis into standard feedback instructions. The bidirectional instruction routing service module is used to parse and route feedback instructions from the intelligent twin client module to achieve closed-loop optimization.

[0016] Compared with the prior art, the present invention has the following beneficial effects: This invention uses a CAM semantic adapter to transform static process data into a dynamic event stream that can drive simulation, solving the time alignment problem of multi-source data from the source, making twin visualization a true reflection of the simulation process, and opening up and activating the entire "process-simulation-twin" link.

[0017] This invention employs incremental synchronization based on simulation logic time and client-side interpolation rendering, which can ensure millimeter-level synchronization between visual performance and logical processes even in complex scenes, achieving a smooth, cinematic experience and realizing high-fidelity, time-sensitive synchronization and visualization.

[0018] This invention assigns actual time, cycle time, and concurrency relationships to static processes in a simulation platform, giving digital twins in-depth engineering analysis capabilities, making it possible to quantitatively analyze production cycle time and capacity bottlenecks.

[0019] This invention constructs a flexible and decoupled intelligent closed loop through abstract feedback instructions and routing mechanisms, which non-destructively feeds back the results of twin interaction or intelligent analysis to the simulation or process system, forming a scalable virtuous cycle of "simulation verification - visualization analysis - decision optimization", while ensuring the stability of the core timing link and the overall resource efficiency of the system. Attached Figure Description

[0020] Figure 1 This is a flowchart of the bidirectional synchronization method for the digital twin platform for flexible manufacturing lines described in this invention.

[0021] Figure 2This is a schematic diagram illustrating the workflow and data conversion of the CAM semantic adapter in this invention.

[0022] Figure 3 This is a schematic diagram illustrating the generation and publication of incremental update messages in this invention. Detailed Implementation

[0023] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0024] Example 1 like Figure 1 As shown, the bidirectional synchronization method for a digital twin platform for flexible manufacturing lines described in this invention includes the following steps: S1 develops and deploys a CAM semantic adapter that connects to the CAM system, semantically parses CAM (Computer-Aided Manufacturing) data, encapsulates it into standardized process events, and then injects it into a general data service platform.

[0025] This invention achieves semantic parsing and encapsulation of CAM (Computer-Aided Manufacturing) data through a CAM semantic adapter. For example... Figure 2 As shown, specifically: The CAM semantic adapter connects to the CAM system and parses the machining data output by the CAM system (such as NC program files). Figure 2 The parsing module extracts process semantic units from CAM machining data, including machining features (such as milling and drilling), toolpath geometric coordinate sequences, process logic sequences, process parameters (spindle speed, feed rate, tool number), and theoretical machining time.

[0026] The standardized process event encapsulation involves encapsulating the parsed process semantic units into standardized process events, such as process start events, process completion events, tool change events, etc., according to the event format and data specifications defined by the general data service platform. Its event payload meets the event payload requirements of standardized process events within the general data service platform, such as carrying complete process semantic information including the target part, process parameters, tool number, and theoretical time consumption. Figure 2 (The encapsulation module in the middle).

[0027] According to the standard interface of the general data service platform (such as OPC UA), the encapsulated semantic unit is injected into the data bus of the general data service platform.

[0028] S2, the simulation platform receives standardized process events from the general data service platform. After simulating the encapsulated standardized process events, the simulation platform generates a time-series data stream. The specific process is as follows: The simulation platform acquires encapsulated, standardized process events by subscribing to process events. Based on the event payloads within these standardized process events, it loads the corresponding toolpath data and process parameters, simultaneously driving its internal high-fidelity virtual production line model (including machine tools, robots, and logistics) to execute a complete machining process simulation. During simulation, the simulation engine dynamically generates a time-series data stream with simulation timestamps for each internal state change, logical event, and motion step. This time-series data stream includes equipment spatial states (coordinates, attitude), discrete events (such as "simulation process start," "collision warning"), material flow snapshots, and the cycle time status of the entire production line.

[0029] The time-series data stream is pushed back to the general data service platform in real time through the data interface of the simulation platform (such as the OPC UA server).

[0030] S3 develops and deploys enhanced digital twin synchronization services, subscribing to and receiving time-series data streams in real time. It compares new time-series data with the previous state snapshot of the corresponding entity, generating incremental update messages. Specifically, when new time-series data is received, it compares the new data with the previous state snapshot of the corresponding entity attribute by attribute, calculating the state differences. Only the changed attributes and their new values, along with their simulation logic timestamps, are encapsulated into a lightweight incremental update message, guiding the digital twin platform on when to update to which state.

[0031] Incremental update messages are prioritized according to business requirements. This ensures that high-priority events such as "simulation collision alarm" and "process timeout" are identified and transmitted immediately. The transport layer schedules the sending queue and network resources based on priority. Incremental messages are delivered to the digital twin platform via a persistent bidirectional connection (such as WebSocket). Figure 3 As shown.

[0032] The core of this step is to maintain and respect the simulation engine's virtual timeline, ensuring that subsequent processing is based on the correct logical timing.

[0033] S4, the digital twin client, drives scene updates based on incremental update messages and the simulation timestamps within those messages. It performs short-term predictions based on the current motion state (velocity, acceleration) of the entity motion model and combines this with interpolation rendering to effectively compensate for network latency and jitter, achieving a smooth visual transition.

[0034] The digital twin client provides a user interface and integrates an intelligent analysis module. When the user interacts (such as clicking "Mark Bottleneck") or the analysis module generates output (such as "Predictive Maintenance Warning"), the output is encapsulated into standardized feedback instructions.

[0035] The digital twin client reveals potential problems in the "space" and "time" dimensions of the physical production line through high-precision kinematic simulation. Specific feedback instruction types include: spatial interference and collision warning, production cycle bottleneck analysis, and logistics scheduling deadlock detection.

[0036] The digital twin client sends feedback instructions back to the general data service platform via a two-way communication link (such as WebSocket). The instruction routing service within the general data service platform accurately distributes these instructions to downstream systems based on their type and target. For example, a "collision warning" or "bottleneck indicator" instruction can be sent to the simulation platform, driving it to adjust the robot's trajectory or logistics strategy in the virtual environment for verification and optimization; or fundamental process issues such as "toolpath interference" can be fed back to the CAM system, triggering optimization prompts for process parameters or toolpaths, thus forming an intelligent closed loop of "simulation verification - visualization analysis - decision optimization." User interaction instructions (such as pause) can be sent to the simulation platform; optimization suggestion instructions can be sent to the simulation platform for verification, or triggered through the platform to prompt the CAM system.

[0037] This invention addresses the bidirectional data and instruction flow in flexible manufacturing lines, from process planning to dynamic simulation to intelligent twinning. It optimizes the transmission link for the time-sequential data flow from the process semantic units of the CAM (Computer-Aided Manufacturing) platform to the simulation platform, the incremental update messages of the digital twin platform, and the standardized feedback instructions. Specifically, it includes: For the data source, an intelligent redundancy filtering strategy is adopted. A dead zone filtering algorithm is applied to high-frequency state data, and it is only allowed to be published when the numerical change exceeds a preset threshold (e.g., position > 1 mm, velocity > 0.5%).

[0038] To avoid the overhead of frequent handshakes, establish persistent and stable bidirectional connections (such as WebSocket), and use lightweight compression techniques, such as GZIP, for batch transmission of ordinary priority data to improve transmission efficiency.

[0039] Establish resilient retransmission and data prioritization strategies. For high-priority data, implement reliable transmission mechanisms such as acknowledgments and retransmissions, and automatically downgrade to transmitting only critical data in the event of network anomalies to ensure the availability of core functions.

[0040] To implement the bidirectional synchronization method for a digital twin platform for flexible manufacturing production lines as described in this invention, the bidirectional synchronization system for a digital twin platform for flexible manufacturing production lines as described in this invention is built on an existing general data service platform, including: The CAM semantic adapter module is used to parse and map the machining data of the CAM system into standardized process semantic units and encapsulate them into standardized process events.

[0041] The timing-driven module of the simulation platform drives the dynamic execution of standardized process events within the simulation platform, generating a timing-driven data stream. This timing-driven data stream is the basis for spatial interference checks and bottleneck analysis.

[0042] The enhanced twin synchronization service module compares the time-series data stream with the previous state snapshot of the corresponding entity attribute by attribute, generates incremental update messages, and synchronizes them to the digital twin platform.

[0043] The intelligent twin client module updates messages incrementally and renders scenes, while converting optimization suggestions generated by user interaction or local intelligent analysis into standard feedback instructions.

[0044] The bidirectional instruction routing service module is used to parse and route feedback instructions from the intelligent twin client module to achieve closed-loop optimization. It distributes control instructions (such as pause) to the simulation platform and optimization suggestions related to process root causes (such as toolpath modification) to the CAM system.

[0045] In the bidirectional synchronization system of the digital twin platform for flexible manufacturing production lines of this invention, the CAM semantic adapter module, the enhanced twin synchronization service module, and the bidirectional instruction routing service module can be deployed on the general data service platform side, or they can be deployed separately. They can interface with the general data platform and various application systems through the general data platform interface, or they can be deployed on the corresponding application system side as needed.

[0046] Example 2 illustrates the specific implementation process of the bidirectional synchronization method of the digital twin heterogeneous platform for flexible manufacturing lines described in this invention, using a flexible manufacturing line for aero-engine disk-type components as an example.

[0047] 1. Application scenarios and hardware / software configuration of the flexible manufacturing line for aero-engine disc components in this embodiment. Production line composition: A demonstration production line including a five-axis CNC machining center, a six-axis collaborative robot, an automated guided vehicle, and a three-dimensional warehouse, used for small-batch, multi-variety processing of parts such as turbine disks and compressor disks.

[0048] Software environment: CAM system: Siemens NX or Mastercam 2024.

[0049] Simulation platform: Visual Components 4.8.

[0050] 3D visualization digital twin platform: Unity 2022 LTS.

[0051] Unified modeling tool: JSON Schema.

[0052] Message middleware: EMQX (MQTT Broker 5.0).

[0053] Hardware and Networks: Industrial servers, workshop data acquisition gateways, gigabit industrial Ethernet.

[0054] 2. The specific implementation process of the bidirectional synchronization method for the digital twin heterogeneous platform for flexible manufacturing production lines described in this invention.

[0055] S1 develops and deploys CAM semantic adapters, which are used to semantically parse CAM (Computer-Aided Manufacturing) data, encapsulate standardized process events, and inject them into a general data service platform.

[0056] A CAM semantic adapter was developed as a separate process. This adapter is configured to monitor the Mastercam post-processor output directory. Specifically, when a new NC program (.NC file) is generated, the CAM semantic adapter starts the parsing engine. The parsing engine not only reads the G / M code but also, in conjunction with the process database, parses the code into a series of standardized process event machining steps. Each step includes: geometric trajectory, spindle speed, feed rate, tool ID, and theoretical time.

[0057] The CAM semantic adapter encapsulates each standardized process event into a structured JSON object, i.e., a standardized process event. Subsequently, through the API exposed by the general data service platform, these standardized process events are sent sequentially in batches to the general data service platform and associated with specific virtual machine bed entities.

[0058] S2, the manufacturing simulation platform obtains encapsulated semantic units by subscribing to process events. Based on the event payloads within these semantic units, it loads the corresponding toolpath data and process parameters, simultaneously driving its internal high-fidelity virtual production line model (including machine tools, robots, and logistics) to execute a complete machining process simulation. During simulation, the simulation engine dynamically generates a time-series data stream with simulation timestamps for each internal state change, logical event, and motion step. This time-series data stream is pushed in real-time to the general data service platform via the simulation platform's data interface (such as an OPC UA server).

[0059] S3, implement enhanced twin synchronization services The enhanced twin synchronization service subscribes to "VC device status" from OPC UA (OPC Unified Architecture). By processing this data stream in real time, the service can fuse discrete states and events (such as device coordinates and process status) during the simulation process into a composite state object containing information such as "machining process," "execution progress," "real-time coordinates," and "cycle time status." For example, when receiving "CAM machining step event (step 2)" and "VC real-time coordinate data" for the same machine tool, they are merged into a composite state object containing "current step," "progress percentage," "real-time coordinates," and "theoretical remaining time."

[0060] The enhanced twin synchronization service maintains the last sent state dictionary for each Unity connection. At each scheduling iteration, the difference between the composite state object and the old dictionary is calculated. Only the changed portions (e.g., "progress percentage changed from 45% to 46%) are serialized into compact binary messages (using MessagePack format), forming incremental update messages. These incremental update messages are then prioritized and placed into high, medium, and low priority queues. A dedicated sending thread prioritizes high-priority queues (e.g., messages containing "alarm" status).

[0061] S4, Implement Unity Smart Twin Client Upgrade For a moving AGV model, a "predicted position" is calculated based on its current speed and the received target point. In Update, Vector3.SmoothDamp is used to make the model move smoothly towards the predicted position, so that the screen remains smooth even if the network frame rate fluctuates.

[0062] When a user right-clicks a "buffer" in a Unity scene and selects "Mark as Bottleneck", a standardized feedback instruction is generated: {"type": "UI_MARK_BOTTLENECK", "entityId": "Buffer_003", "timestamp": ...}.

[0063] The intelligent twin client is integrated into a lightweight SPF-LSTM model in Unity. It can analyze real-time vibration data streams, predict spindle anomalies, and generate standardized feedback commands: {"type": "PREDICTIVE_ALERT", "entityId": "Spindle_01", "alertCode":"VIBRATION_HIGH", "confidence": 0.88}.

[0064] Standardized feedback instructions are immediately sent to the general data service platform via the WebSocket connection's return channel.

[0065] S5 implements general data service-side command routing and response. Deploy a bidirectional command routing service within the general data platform. This service listens to the WebSocket postback channel. Upon receiving the UI_MARK_BOTTLENECK command, it writes a variable to the Visual Components simulation platform via an OPC UA write operation, according to the routing rules (configuration table). This triggers an internal VC script, highlighting the corresponding cache model in the simulation interface.

[0066] Upon receiving the PREDICTIVE_ALERT command, two routes are executed simultaneously: first, the VC is notified via OPC UA to record a warning in the simulation log; second, the adapter displays a prompt box in the Mastercam plugin interface by calling the callback interface provided by the CAM semantic adapter.

[0067] 3. Implementation effect verification To quantitatively evaluate the effectiveness of this invention, the inventors constructed a test environment matching the flexible manufacturing demonstration production line of an aero-engine manufacturing company. The proposed solution (experimental group) was compared with two typical existing integrated solutions in a continuous 72-hour test. The comparison scheme is designed as follows: Comparison of key performance indicators: End-to-end event synchronization delay: For the processing start command, the average delay of the present invention is 122 ms; the average delay of the traditional solution is 142 ms.

[0068] Network bandwidth usage: Under the condition of simulating the full-speed operation of the production line, the average bandwidth usage of the present invention is 27Mbps; the average bandwidth usage of the traditional solution is 32Mbps.

[0069] Closed-loop response capability: This invention successfully achieves a fully automated closed loop from "twin prediction alarm" to "simulation model adjustment" and "CAM interface prompts," with an average response time of less than 1.5 seconds. Traditional solutions cannot achieve this automated reverse link.

[0070] The methods and systems described in this invention are not limited to the specific software brands, communication protocols, or production line layouts mentioned in the above embodiments. Any technical solution that utilizes the core idea of ​​"using a general data platform as a data interaction hub and achieving bidirectional closed-loop synchronization between multi-source heterogeneous platforms through semantic adaptation extension and synchronization service enhancement" falls within the scope of protection intended by this invention.

[0071] This invention addresses the existing general-purpose data service platform architecture and achieves the following core improvements: It has opened up and activated the entire "process-simulation-twin" chain: through the CAM semantic adapter, static process data is transformed into a dynamic event stream that can drive simulation, solving the time alignment problem of multi-source data from the source, and making twin visualization a true reflection of the simulation process.

[0072] It achieves high-fidelity, time-sensitive synchronization and visualization: by using incremental synchronization based on simulation logic time and client-side interpolation rendering, it can still ensure millimeter-level synchronization between visual performance and logical process in complex scenes, achieving a smooth, cinematic experience.

[0073] Digital twins are endowed with deep engineering analysis capabilities: the simulation platform gives static processes actual time, cycle time and concurrency relationships, and the generated time-series data makes it possible to quantitatively analyze production cycle time and capacity bottlenecks.

[0074] A flexible and decoupled intelligent closed loop was constructed: through abstract feedback instructions and routing mechanisms, the results of twin interaction or intelligent analysis are fed back to the simulation or process system without loss, forming a scalable virtuous cycle of "simulation verification - visualization analysis - decision optimization", while ensuring the stability of the core timing link and the overall resource efficiency of the system.

Claims

1. A bidirectional synchronization method for a digital twin platform for flexible manufacturing production lines, characterized in that, Includes the following steps: S1, develop and deploy CAM semantic adapters to semantically parse CAM data and encapsulate it into standardized process events before injecting it into a general data service platform; S2, the simulation platform simulates the packaged standardized process events, generates a time-series data stream, and injects it into the general data service platform; S3 is used to develop and deploy enhanced twin synchronization services, subscribe to and receive time-series data streams in real time, compare new time-series data with the previous state snapshot of the corresponding entity, and generate incremental update messages. S4, the 3D twin client drives scene updates based on incremental update messages, outputs standardized feedback commands through user interaction or intelligent analysis, and sends them back to the general data service platform through a two-way communication link; S5, the general data service platform, distributes feedback instructions to downstream systems based on the type and target of the instructions.

2. The bidirectional synchronization method for a digital twin platform for flexible manufacturing lines according to claim 1, characterized in that: The time-series data stream consists of state changes, logical events, and motion steps with simulation timestamps, generated by the simulation platform driving the virtual production line model based on standardized process events.

3. The bidirectional synchronization method for a digital twin platform for flexible manufacturing lines according to claim 1, characterized in that: The incremental update message is the state difference value between the current time-series data stream and the previous state snapshot of the corresponding entity.

4. The bidirectional synchronization method for a digital twin platform for flexible manufacturing lines according to claim 1, characterized in that: An intelligent redundancy filtering strategy is adopted, and data is only released to the general data service platform when the data changes exceed a preset threshold.

5. The bidirectional synchronization method for a digital twin platform for flexible manufacturing lines according to claim 1, characterized in that: The feedback instruction types include control instructions and optimization suggestions for process root causes; the control instructions are distributed to the simulation platform; and the optimization suggestions for process root causes are distributed to the CAM system.

6. A two-way synchronization system for a digital twin platform for flexible manufacturing production lines, based on a general data service platform, characterized in that: include: The CAM semantic adapter module is used to parse and map the machining data of the CAM system into standardized process semantic units and encapsulate them into standardized process events. The timing-driven module of the simulation platform drives the dynamic execution of standardized process events within the simulation platform, generating a timing-driven data stream. The enhanced twin synchronization service module compares the time-series data stream with the previous state snapshot of the corresponding entity attribute by attribute, generates incremental update messages, and synchronizes them to the 3D twin platform. The intelligent twin client module renders the scene based on incremental update messages, and converts optimization suggestions generated by user interaction or local intelligent analysis into standard feedback instructions. The bidirectional instruction routing service module is used to parse and route feedback instructions from the intelligent twin client module to achieve closed-loop optimization.

Images