Search

CN-121979123-A - Intelligent manufacturing digital twin real-time synchronization system

CN121979123ACN 121979123 ACN121979123 ACN 121979123ACN-121979123-A

Abstract

The invention relates to the technical field of intelligent manufacturing, and discloses an intelligent manufacturing digital twin real-time synchronization system. The invention aims to solve the problems of high response delay, large bandwidth pressure and poor model suitability of the traditional digital twin system. According to the invention, the model is decoupled from the working condition, the global model iteration is driven by using edge trial and error, the network load is greatly reduced, and the control precision of the full life cycle is improved. The system corrects general parameters by combining the real-time working condition through the edge node, performs closed-loop debugging and spot check monitoring, uploads mixed data of trial and error and spot check, and the cloud uses an uploading data synchronization twin body to update the global library and optimize the core model through generalized restoration so as to directionally issue a dedicated increment update package.

Inventors

  • YANG JIAN

Assignees

  • 云尖信息技术股份有限公司

Dates

Publication Date
20260505
Application Date
20251215

Claims (10)

  1. 1. The intelligent manufacturing digital twin real-time synchronization system is characterized by comprising edge computing nodes deployed on a physical device side and a cloud server, and the intelligent manufacturing digital twin real-time synchronization system is cooperated with the cloud server through a bidirectional encryption communication link; The edge computing node is used for receiving general core parameters issued by a cloud, carrying out localized correction by combining with equipment working condition parameters acquired in real time to generate initial process parameters, executing local closed loop debugging based on quality difference of finished products in a trial production stage to generate and store trial-error data, executing spot check monitoring in a stable production stage and recording key data, and packing and uploading a mixed data set containing the trial-error data and the spot check data by combining with a digital watermark; The cloud server is used for verifying the digital watermark of the uploaded data and analyzing the working condition data to synchronize the digital twin copy corresponding to the edge computing node, carrying out generalized restoration on the uploaded process parameters by utilizing the working condition data, inverting the parameters under the specific working condition into standard general parameters and updating the global parameter mapping library, optimizing the general core parameters based on the updated global parameter mapping library, generating an updating packet containing the special core parameter correction value and the increment mapping data, and transmitting the updating packet to the edge computing node.
  2. 2. The intelligent manufactured digital twin real-time synchronization system according to claim 1, wherein the process of performing the localization correction by the edge computing node to generate an initial process parameter specifically comprises activating a local state monitoring module, reading current working condition parameters of a physical device through a sensor network, wherein the working condition parameters comprise wear degree of mechanical parts, ageing coefficient of an actuating mechanism and current environmental thermodynamic state, calling a locally stored general core parameter, wherein the general core parameter comprises a physical model reference value describing the operation of the device in an ideal standard state, taking the working condition parameter as a correction variable into a physical model of the general core parameter, calculating a parameter compensation amount required for counteracting performance degradation of the device, superposing the parameter compensation amount onto a standard process instruction derived from the general core parameter, synthesizing the initial process parameter adapted to the current device health state, and mapping the initial process parameter into a bottom layer execution code of a device controller.
  3. 3. The intelligently manufactured digital twin real-time synchronization system according to claim 2, wherein the process of the edge computing node performing the local closed loop debugging during the commissioning phase specifically comprises: after the initial technological parameters are utilized to finish the first trial production, physical dimension parameters of actual finished products are collected through detection equipment, and difference values between the physical dimension parameters and preset target finished product parameters are calculated; If the difference value exceeds a preset quality convergence threshold, judging that the current process parameter does not reach the standard, performing fine adjustment correction on the previous process parameter by utilizing parameter deviation in a local calculation unit, generating a new test process parameter, and then performing next trial production; In each round of the iterative loop, whether the trial production result reaches the standard or not, combining the currently used test process parameter, the corresponding actual finished product physical dimension parameter and the current equipment working condition parameter into a complete trial-error record, marking the complete trial-error record as high-value sample data, and storing the high-value sample data into a local temporary parameter library until the difference value is smaller than the quality ending threshold value, and completing debugging and locking the final production process parameter.
  4. 4. The intelligent manufactured digital twin real time synchronization system according to claim 3, wherein the process of performing the spot check monitoring by the edge computing node during the stable production phase comprises: After locking final production process parameters and starting continuous production, triggering a sampling instruction according to a preset time interval or a preset yield step, collecting physical parameters of a finished product in a current spot check period and synchronous equipment working condition parameters, and comparing the physical parameters obtained by spot check with the target finished product parameters; If the comparison result shows that the deviation is in the allowable range, keeping the current production process parameters unchanged, not executing any parameter adjustment operation, and only storing the process parameters of the spot check time period, the physical parameters of the finished product and the working condition parameters of the equipment as a simplified production monitoring record into a local storage area; if the comparison result shows that the deviation exceeds the allowable range, the stable production state is terminated and the local closed loop debugging process is restarted to correct the process parameters.
  5. 5. The intelligently manufactured digital twin real-time synchronization system according to claim 1, wherein the process of the edge computing node for packing and uploading the error test data and the spot check data specifically comprises: After the completion of the current production batch is detected, extracting a mixed data set consisting of the trial-and-error data and the spot check data from a local storage area, and performing denoising cleaning on the mixed data set; performing lossless compression encoding to completely preserve boundary features of parameter adjustment for trial-and-error data in the hybrid dataset marked as a debugging stage; after fixed-point differential compression of the cleaned data is completed, a non-tamperable edge digital watermark is calculated and generated by combining a local private key of an edge node, a hash abstract of a data packet, a unique equipment identity, a current timestamp and a data serial number, and is embedded into a metadata segment of the data packet, and then the data packet with the edge digital watermark is sent to a cloud server through an uplink channel of a communication link.
  6. 6. The intelligently manufactured digital twin real-time synchronization system according to claim 1, wherein the cloud server performs the process of verifying the digital watermark of the uploaded data and parsing the operating mode data to synchronize the digital twin copy specifically comprises: after receiving a data packet uploaded by an edge computing node, firstly decrypting and verifying the edge digital watermark in the data packet by utilizing a prestored public key corresponding to the dialogue node, and checking the continuity of the serial number of the data packet; The method comprises the steps of verifying, peeling off watermarks after passing, decompressing a data packet, analyzing equipment working condition parameters, test error data and spot check data contained in the data packet, positioning and updating state attributes of digital twin copies of corresponding equipment in a cloud database by utilizing the analyzed latest equipment working condition parameters, and finishing synchronization of digital characteristics and actual abrasion and aging states of physical entities; And if the verification fails, triggering a safety alarm and terminating the flow.
  7. 7. The intelligent manufactured digital twin real-time synchronization system according to claim 6, wherein the cloud server executing the process of generalized restoration and updating the global parameter mapping library specifically comprises calling a digital twin copy with updated state attributes, extracting the equipment wear offset value and the aging coefficient recorded therein; Calculating the difference between the theoretical process parameter value and the uploaded actual process parameter, as a process compensation amount caused by the current working condition, reducing the actual process parameter into a general process parameter after the influence of the stripping working condition; in the process, test error data which are generated in the debugging stage and have the quality of finished products which are not up to standard are screened out, and the corresponding general process parameters are marked as negative sample boundaries in a parameter library and are used for limiting the feasible domain range of the follow-up core parameter optimization.
  8. 8. The intelligent manufactured digital twin real-time synchronization system according to claim 7, wherein the cloud server performs the process of optimizing the general core parameters and generating the update package specifically comprises: Based on the updated global process and finished product parameter mapping library, recalculating coefficient values of the general core parameters in a parameter feasible domain range until the calculated predicted finished product quality corresponding to the process parameters and the actual recorded finished product quality errors in the library are smaller than deviation threshold values after compensating amounts of any recorded working conditions of the general core parameters in the superposition library, locking the optimized general core parameters, and reading latest equipment working condition parameters aiming at edge calculation nodes of the uploaded data; The method comprises the steps of optimizing a universal core parameter, combining the latest equipment working condition parameter of a node, calculating a special core parameter correction value required by the next stage of a dialogue node in advance, comparing a parameter library version of the dialogue node in last synchronization, extracting a standard mapping pair which is newly added in the interaction period and is verified to be qualified from a global mapping library to be used as incremental mapping data, and packaging the special core parameter correction value and the incremental mapping data to generate an update package.
  9. 9. The intelligent manufactured digital twin real-time synchronization system according to claim 8, wherein the process of the cloud server for directing the update package down specifically comprises: After generating an update package, combining a current update time stamp and a model version number, calculating by using a cloud private key to generate a cloud digital watermark, and embedding the cloud digital watermark into the head of the update package; Establishing a downlink transmission channel only pointing to the edge computing node of the conversation, and single-point transmitting an update packet carrying the watermark to the edge computing node of the conversation; And after receiving the update package, the edge computing node verifies the watermark validity by utilizing the cloud public key, namely, the exclusive core parameter correction value is applied to the local control logic after verification is passed, and the increment mapping data is merged into a local parameter mapping library to complete single closed loop synchronization.
  10. 10. The intelligently manufactured digital twin real-time synchronization system according to claim 1, wherein the communication link between the edge computing node and the cloud server employs a dual channel priority routing strategy: Configuring a high-bandwidth uplink channel, which is specially used for transmitting a mixed data set which is sent by an edge computing node and contains high-capacity error test data and spot check data; A low-delay downlink channel is configured, and is endowed with higher network scheduling priority than an uplink channel, and the low-delay downlink channel is specially used for transmitting an update packet which is issued by a cloud server and contains a core parameter correction value and incremental mapping data so as to ensure the real-time arrival of a control instruction.

Description

Intelligent manufacturing digital twin real-time synchronization system Technical Field The invention relates to the technical field of intelligent manufacturing, in particular to a digital twin real-time synchronization system for intelligent manufacturing. Background Can be manufactured as a product of deep fusion of a new generation of information technology and an advanced manufacturing technology, and the production mode and the value chain form of the global industry are deeply changed. The digital twin technology is used as one of key enabling technologies for intelligent manufacturing, and the digital simulation and full life cycle management of the whole production elements and the whole process are realized by constructing a digital model which is mapped with a physical entity in real time in a virtual network space. Currently, digital twin application based on industrial Internet has been widely permeated into the fields of aerospace, precision machining, automatic production lines and the like, so that the optimal configuration of production resources is effectively promoted, the flexibility level and collaborative operation capability of a manufacturing system are improved, and the digital twin application becomes an important engine for pushing the manufacturing industry to realize digital, networked and intelligent transformation. However, in the face of large-scale, high-frequency industrial real-time control scenarios, there are still significant limitations to the existing digital twinning technology architecture. Firstly, the traditional centralized data processing mode requires that mass equipment data is uploaded to a cloud end in full quantity, so that network bandwidth is overloaded, and data transmission and processing delay often cannot meet the severe requirement of precision manufacturing on millisecond-level response. In addition, the conventional modeling method generally adopts a standardized general model to drive the equipment of the same type, ignores individual working condition differences of physical entities caused by mechanical abrasion, component aging or environmental change in the long-term operation process, and leads to deviation of control instructions when the general parameters based on the ideal state are directly applied to the equipment with performance degradation, so that the consistency of the quality of finished products is difficult to ensure, and even the operation failure of the equipment is possibly caused. Disclosure of Invention The present invention has been made in view of the above-described problems occurring in the prior art. Therefore, the invention provides an intelligently manufactured digital twin real-time synchronization system, which solves the problems of high response delay, high bandwidth pressure and poor model suitability of the traditional digital twin system. In order to solve the technical problems, the invention provides the following technical scheme: the invention provides an intelligent manufactured digital twin real-time synchronization system, which consists of an edge computing node and a cloud server, wherein the edge computing node is deployed at a physical device side, and the edge computing node and the cloud server cooperate to work through a bidirectional encryption communication link; The edge computing node is used for receiving general core parameters issued by a cloud, carrying out localized correction by combining with equipment working condition parameters acquired in real time to generate initial process parameters, executing local closed loop debugging based on quality difference of finished products in a trial production stage to generate and store trial-error data, executing spot check monitoring in a stable production stage and recording key data, and packing and uploading a mixed data set containing the trial-error data and the spot check data by combining with a digital watermark; The cloud server is used for verifying the digital watermark of the uploaded data and analyzing the working condition data to synchronize the digital twin copy corresponding to the edge computing node, carrying out generalized restoration on the uploaded process parameters by utilizing the working condition data, inverting the parameters under the specific working condition into standard general parameters and updating the global parameter mapping library, optimizing the general core parameters based on the updated global parameter mapping library, generating an updating packet containing the special core parameter correction value and the increment mapping data, and transmitting the updating packet to the edge computing node. The method comprises the steps of firstly activating a local state monitoring module, reading current working condition parameters of physical equipment through a sensor network, wherein the working condition parameters comprise mechanical component abrasion degree, an execution mechanism aging coefficien