Detection laboratory total factor digital twinborn model construction method and intelligent operation and maintenance system

Abstract

The invention discloses a detection laboratory total factor digital twinning model construction method and an intelligent operation and maintenance system, and relates to the field of digital twinning, and a total factor digital twinning model comprises a detection laboratory physical object model, a detection laboratory virtual model, a total factor data model, a data interaction model, a learning decision model and an operation and maintenance management model. Real-time mapping is kept between the physical object model and the virtual model of the detection laboratory; the total element data model is used for collecting, inputting, processing and updating multi-source heterogeneous data in the digital twinborn model of the detection laboratory; the data interaction model carries out fusion transmission on real-time data among the models; and meanwhile, the learning decision-making model performs autonomous decision making and diagnosis, and functions are integrated and displayed through the operation and maintenance management model, so that the intelligent operation and maintenance level of the detection laboratory is further improved. The whole detection process of the detection laboratory is digitalized, the communication cost of the detection test process is saved, the manpower and time cost is saved, and the detection efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of digital twin technology, and more specifically, to a method for constructing a full-factor digital twin model of a testing laboratory and an intelligent operation and maintenance system. Background Art

[0002] A testing laboratory is a place where one or more characteristics or performances of a given product, material, equipment, organism, physical phenomenon, process or service are determined according to prescribed procedures. Testing laboratories play an extremely important role in the production and research and development process of the equipment manufacturing industry. Test results and test reports are the key data that large equipment manufacturing companies focus on in their goals of improving production processes, improving product performance, and reducing production costs.

[0003] During the operation of the testing laboratory, the biggest obstacle to achieving low-cost operation and maintenance management of the testing laboratory is the labor-intensive, time-consuming and labor-intensive work of statistical sorting of testing data. It is difficult to fully grasp the real-time status of on-site personnel, equipment, materials, testing methods, testing environment, etc., and the laboratory's real-time testing data and problem feedback and corresponding processing are not timely enough. Test data statistics, test process management and problem handling decisions mainly rely on the experience of staff. The working efficiency of the testing laboratory is low, and it is impossible to quickly and accurately analyze and mine a large amount of test data. Summary of the invention

[0004] In view of this, the present invention provides a full-factor digital twin model and operation and maintenance system of a testing laboratory operation and maintenance system, which are used to solve the problem of low efficiency of operation and maintenance management of existing testing laboratories.

[0005] To achieve the above objectives, the proposed solution is as follows: A method for constructing a full-factor digital twin model of a testing laboratory, comprising: Step 1: Analyze the physical objects and data collection objects of the detection laboratory and build a physical object model of the detection laboratory. EO is the physical object of the testing laboratory, and DCO is the data collection object of the testing laboratory; the physical object of the testing laboratory includes the laboratory layout, testing equipment, tested samples, testing personnel, and testing conditions; the data collection object of the testing laboratory includes data collection equipment and data transceiver modules. The data collection equipment includes laboratory data collection equipment for testing related sensors and PLCs, and the data transceiver modules include data collection gateways and industrial switches; Step 2: Based on the physical object model of the test laboratory, a virtual model of the test laboratory based on digital twin is constructed. GM is the geometric model of the testing laboratory; RM is the rule model of the testing laboratory; FM is the process model of the testing laboratory; the laboratory geometric model of the test environment and testing equipment is constructed by means of 3D scanning modeling, point cloud data processing and mesh surface rendering; the laboratory process model is established by combing the mechanism coupling relationship between the basic information of the testing laboratory and the testing process information; the laboratory rule model is constructed through the laboratory geometric model, laboratory process model and the operation and maintenance system functions of multi-equipment mechanism linkage and cross-regional test data sharing; Step 3: Based on the test laboratory data sets of each model, construct the full-factor data model of the test laboratory, where the full-factor data model is ; a test lab environment configuration data set generated when running the test lab physical object model; The basic data of the test equipment, the simulation data of the geometric model and the test result data of the process model when running the virtual model of the test laboratory; and They are respectively the operation and maintenance status data and operation and maintenance knowledge data generated when running the operation and maintenance management system and function display; In order to run the learning decision model, the above multi-source data is autonomously analyzed, learned and decided to form a decision optimization data set; Step 4: According to the operation and maintenance management tasks of the testing laboratory, build an operation and maintenance management model, where the operation and maintenance management model is , TS is the technical service module, UT is the user service module, and FA is the functional application module; The technical service module is a collection of technical services required for the function operation. The technical service modules are: For the model service module, For the data service module, It is a comprehensive service module; The user service module is a service support built around the full-factor resources of man-machine-material-method-environment-test-audit in the whole process of testing; the functional application module is used to integrate the functional services required in the operation and maintenance management of the testing laboratory; Step 5: Construct a learning decision model to analyze the status of the testing equipment in the testing laboratory based on the real-time data received by the data interaction model. The learning decision model is , It is the execution control based on the action strategy selected according to the detection task and laboratory environment information. The reward and punishment mechanism includes reward function and penalty function. Set conditions for the algorithm; Step 6: Based on the constructed physical object model of the testing laboratory, the virtual model of the testing laboratory, the operation and maintenance management model, the learning decision model and the full-factor data model, the data interaction relationship between the various models of the testing laboratory is analyzed, and the data interaction model of the testing laboratory is constructed. , , and They are the real-time data interaction between the physical object model of the testing laboratory and the full-factor data model, the virtual model of the testing laboratory and the operation and maintenance management model; and They are the real-time data interaction between the virtual model of the testing laboratory and the full-factor data model and the operation and maintenance management model respectively; To provide real-time data interaction between the operation and maintenance management model and the full-factor data model; Real-time data interaction between the full-factor data model and the learning decision model; Step 7: Based on the constructed models of the testing laboratory, analyze the correlation between the testing process and all the elements of the testing process, and build a digital twin model of the elements, so that the staff can build an intelligent operation and maintenance system for the testing laboratory based on the full-element digital twin model. The full-element digital twin model is .

[0006] Preferably, in step 2, the process model is used to control the docking process of the overall business process of the testing laboratory; the rule model is used to describe and characterize the test information and testing process of the sample to be tested; The basic information for building the process model includes: laboratory environment, test conditions, test equipment and operators; the test process information includes test entrustment, test execution and report release.

[0007] Preferably, in step 3, the test laboratory environment configuration data set includes: test equipment operating status, test personnel configuration, test cycle, laboratory power supply, ventilation and temperature and humidity environment; The operation and maintenance status data generated during the operation of the operation and maintenance management system includes data type conversion, preprocessing, classification and integration; Operation and maintenance knowledge data includes inspection records, repair and maintenance records, knowledge and experience of testing experts, testing equipment manuals and laboratory management plans.

[0008] Preferably, the process of analyzing the state of the testing equipment in the testing laboratory by the learning decision model using the real-time data received by the data interaction model includes: After the learning decision model receives the real-time data transmitted by the data interaction model, it matches the real-time data from different sources with the algorithms in the preset algorithm set, and performs fault diagnosis, anomaly detection, remaining service life prediction and multi-objective optimization on the testing equipment in the testing laboratory based on the matched algorithms.

[0009] Preferably, in step 5, the execution control of the action strategy selected according to the detection task and the laboratory environment information includes: installation of the sample to be tested, parameter setting, performance testing, environmental adaptability testing, data collection and analysis, and report preparation; The reward function in the reward and punishment mechanism is used to motivate the operator to select the current action strategy signal, and the penalty function is used to adjust the operator's current action strategy signal. The reward function includes action reward, position reward and distance reward, and the penalty function includes collision penalty, intra-region penalty and time penalty. The algorithm setting conditions are the conditions that need to be set to complete learning and decision-making training, including algorithm network structure design, training parameter setting, and round termination condition setting.

[0010] Preferably, in step seven, all the testing elements include testing personnel - testing equipment - testing samples - testing methods - testing environment - testing results - report review; the entire testing process includes test environment configuration, testing personnel operation, laboratory operation and maintenance management, system intelligent decision-making and test report generation.

[0011] An intelligent operation and maintenance system for a testing laboratory is constructed based on the full-factor digital twin model constructed by the aforementioned construction method, and the system architecture includes: a testing laboratory, a system support layer, a twin model layer, and an application display layer; The system functional modules include laboratory operation module, laboratory management module, test equipment management module and test data analysis module.

[0012] It can be seen from the above technical scheme that the present invention provides a method for constructing a full-factor digital twin model of a detection laboratory and an intelligent operation and maintenance system. The constructed full-factor digital twin model includes: a physical object model of the detection laboratory, a virtual model of the detection laboratory, a full-factor data model, a data interaction model, a learning decision model and an operation and maintenance management model. The full-factor data model is used to collect, input, process and update multi-source heterogeneous data in the full-factor digital twin model; the data interaction model integrates and transmits real-time data between the models in the full-factor digital twin model through a preset communication protocol. The present invention realizes the digitization of the detection process. Through the real-time mapping of the physical object model of the detection laboratory and the virtual model of the detection laboratory, the communication cost of the detection test process is saved, the manpower and time cost is saved, and the detection efficiency is improved. At the same time, the introduction of machine learning algorithms for autonomous decision-making and diagnosis further improves the level of intelligent operation and maintenance management and control of the detection laboratory.

[0013] The digital system can effectively avoid human errors and negligence, and further improve the accuracy of test reports. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0015] Figure 1 A schematic diagram of the relationship between the various models of a full-factor data twin model of a testing laboratory provided in an embodiment of the present invention; Figure 2 A schematic diagram of the process of constructing a full-factor data model provided by an embodiment of the present invention; Figure 3 A schematic diagram of all-factor resource data provided by an embodiment of the present invention; Figure 4 A full flow chart of detection provided by an embodiment of the present invention; Figure 5 A flow chart for constructing a learning decision model provided by an embodiment of the present invention; Figure 6 An architecture diagram of the detection laboratory operation and maintenance system provided by an embodiment of the present invention; Figure 7 A flow chart for constructing a detection laboratory operation and maintenance system provided by an embodiment of the present invention; Figure 8 A functional schematic diagram of a detection laboratory operation and maintenance system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] First, combine Figure 1 The full-factor digital twin model of the detection laboratory operation and maintenance system provided by the embodiment of the present invention is introduced. The present invention constructs a full-factor digital twin model for the detection laboratory operation and maintenance system based on the full-factor detection data and multi-layer twin model analysis for the detection laboratory's entire detection process. : .

[0018] Digital twin model of all elements of the testing laboratory operation and maintenance system It includes the physical object model PM of the testing laboratory, the virtual model VM of the testing laboratory, the full-factor data model DD, the data interaction model CN, the learning decision model LD and the operation and maintenance management model Ss. The relationship between each model is as follows: Figure 1 shown.

[0019] The physical object model PM and the virtual model VM of the testing laboratory are the digital twin models of all elements of the testing laboratory. The general basic model of .

[0020] The full-factor data model DD is the full-factor digital twin model of the testing laboratory The core driving model is responsible for collecting, inputting, processing and updating the full-factor digital twin model. The database is used to store the data in a unified manner, and different data management systems are established for different models such as the physical object model PM of the testing laboratory, the virtual model VM of the testing laboratory, and the operation and maintenance management model Ss, to drive the full-factor digital twin model in real time. The models in the program run dynamically.

[0021] The data interaction model CN ​​transmits real-time data in the full-factor digital twin model through a preset communication protocol. The data interaction model uses a series of communication protocols and communication equipment to integrate all real-time data such as physical perception data, simulation data, and control command signals in the full-factor digital twin model. The models are fused and transmitted to each other, so that the physical object model PM of the detection laboratory and the virtual model VM of the detection laboratory can maintain real-time mapping.

[0022] The learning decision model LD is based on the full-factor data model DD. The learning decision model LD analyzes the status of the detection equipment in the detection laboratory through the real-time data received by the data interaction model CN, obtains the analysis results, and feeds the analysis results back to the operation and maintenance management model Ss for display. After receiving the real-time data transmitted by the data interaction model CN, the learning decision model LD matches the real-time data from different sources with the algorithms in the preset algorithm set, and analyzes the status of the detection equipment in the detection laboratory based on the matched algorithms. The preset algorithm set can be a set of intelligent operation and maintenance decision algorithms formed by calling a series of machine learning algorithms such as gradient boosting, denoising autoencoder, support vector machine, deep belief network, etc. Use powerful computing resources for offline training and model deployment of intelligent operation and maintenance algorithms. Perform fault diagnosis, anomaly detection, remaining service life prediction, and multi-objective optimization on laboratory detection equipment through real-time data received by the data interaction model CN.

[0023] The operation and maintenance management model Ss, based on the existing model technology support and user terminal needs, realizes the operation and maintenance management of the testing laboratory / test equipment, test data analysis and other functional applications through the big data storage / computation / mining / analysis technology in the full-factor data model DD and the predictive maintenance technology of the testing equipment based on artificial intelligence in the learning decision model LD. The laboratory staff can adjust and optimize the management strategy according to the analysis results displayed by the operation and maintenance management model Ss, thus realizing the intelligent operation and maintenance of the testing laboratory.

[0024] It can be seen from the above technical solutions that the embodiment of the present invention provides a method for constructing a full-factor digital twin model of a detection laboratory and an intelligent operation and maintenance system. The constructed full-factor digital twin model includes: a physical object model of the detection laboratory, a virtual model of the detection laboratory, a full-factor data model, a data interaction model, a learning decision model and an operation and maintenance management model. The full-factor data model is used to collect, input, process and update multi-source heterogeneous data in the full-factor digital twin model; the data interaction model integrates and transmits real-time data between the models in the full-factor digital twin model through a preset communication protocol. The embodiment of the present invention realizes the digitization of the detection process. Through the real-time mapping of the physical object model of the detection laboratory and the virtual model of the detection laboratory, the communication cost of the detection test process is saved, the manpower and time cost is saved, and the detection efficiency is improved. At the same time, the introduction of machine learning algorithms for autonomous decision-making and diagnosis further improves the intelligent operation and maintenance management level of the detection laboratory.

[0025] Next, the embodiment of the present invention introduces a method for constructing a full-factor digital twin model of a detection laboratory, and the process is as follows: Step 1: For the testing laboratory, specifically analyze the testing laboratory entity objects and data collection objects, and build the testing laboratory physical object model.

[0026] Specifically, the physical object model PM is the basis of the full-factor digital twin model, and is the main part of the operation and maintenance system and a collection of various management elements. It mainly includes the detection laboratory entity objects (EO, Enity Objects) and data collection objects (DCO, Data Collection Objects). The detection laboratory physical object model is The physical objects of the testing laboratory include the laboratory layout, testing equipment, tested samples, operators and other elements; the data collection objects of the testing laboratory include testing-related sensors, PLC and other laboratory data collection equipment and data collection gateways, industrial switches and other data transceiver modules.

[0027] Step 2: Based on the physical object model of the testing laboratory, build a virtual model of the testing laboratory based on digital twin.

[0028] Specifically, the virtual model VM of the testing laboratory is the engine of the full-factor digital twin model. It is the real-time mapping of the physical object model of the testing laboratory in the virtual scene. It mainly includes the laboratory's geometry model (Geometry Model, GM), rule model (Rule Model, RM) and process model (Flow Model, FM). .

[0029] The test environment and test equipment laboratory geometry model is constructed by 3D scanning modeling, point cloud data processing and mesh patch rendering. For example, the environment scanning method is used to construct the test laboratory equipment and environment geometry parameters and physical motion characteristics. Based on the entity objects in the test laboratory physical object model, a 3D simulation model is constructed through 3D rendering and model lightweight engine to achieve the construction, rendering and lightweight optimization of the test environment and test equipment laboratory geometry simulation model.

[0030] By combing the mechanism coupling relationship between the basic information of the testing laboratory and the testing process information, a laboratory process model is established. Among them, the basic information includes the laboratory environment, test conditions, testing equipment, operators, etc.; the testing process information includes test entrustment, test execution and report release, etc. By combing the coupling relationship and related mechanism formulas between the basic information and the testing process information, the process model is used to control the overall business process docking process of the testing laboratory, solve the problems of resource scheduling and report release in the testing process, realize the standardization and visualization of the business process of the testing laboratory, and improve the operation and maintenance efficiency.

[0031] The laboratory rule model is constructed through a series of operation and maintenance system functions such as the geometric model of the test laboratory, the laboratory process model, multi-equipment mechanism linkage, and cross-regional test data sharing. The rule model is used to fully describe and characterize the test information and test process of the sample to be tested.

[0032] Step 3: Based on the test laboratory data sets of each model, construct a full-factor data model of the test laboratory.

[0033] Specifically, the full-factor data model DD is as follows Figure 2 As shown in Figure 1, the collection and driving interaction of the full-factor data model DD is the driving force of the real-world testing laboratory operation and maintenance management system. The full-factor data model DD is aimed at Figure 3 The full-factor data of the inspection process of "man-machine-material-method-environment-test-audit" is shown in the figure. The full-factor data model is ,in, The test laboratory environment configuration data set generated when running the test laboratory physical object model includes the laboratory site, environment, test condition template, test equipment operation status and personnel configuration, test cycle, laboratory power supply, ventilation, temperature and humidity environment, various facilities and their corresponding lighting, ventilation, water, electricity, gas and other operation data; The basic data of the test equipment, the simulation data of the geometric model and the test result data of the process model when running the virtual model of the test laboratory; and They are respectively the operation and maintenance status data and operation and maintenance knowledge data generated when running the operation and maintenance management model; In order to run the learning decision model, the above multi-source data is autonomously analyzed, learned and decided to form a decision optimization data set.

[0034] Data from the virtual model of the testing laboratory , including basic equipment data such as test equipment number, name, model, manufacturer, quantity, unit price, purchase date, and laboratory ; Geometric model simulation data of equipment such as test equipment startup, shutdown, failure, equipment detection results, etc. ; Supported test items, test standards, test methods, operating parameters and test results and other process model test result data .

[0035] Operation and maintenance status data Including data type conversion, preprocessing, classification, integration, etc. generated during the operation and maintenance management system and function display process; operation and maintenance knowledge data Including inspection records, repair and maintenance records, knowledge and experience of testing experts, testing equipment manuals, spare parts management and replacement records and laboratory management plans, etc.

[0036] The learning decision model is run to complement and integrate the above multi-source data and combine it with the random forest algorithm for autonomous analysis, learning, and decision-making, so as to optimize the prediction of production scheduling management plans such as equipment maintenance and personnel arrangement in each testing laboratory and form a decision optimization data set. The learning decision model LD combines historical statistical data, real-time monitoring data, and process twin data for autonomous analysis, learning, decision-making, and reasoning, so as to optimize the prediction of production scheduling management plans such as equipment maintenance and personnel arrangement in each testing laboratory, analyze the data such as the number of contracts, working hours, and contract amounts over the years, improve the utilization rate of testing equipment, and maximize the testing capacity.

[0037] Step 4: Build an operation and maintenance management model based on the operation and maintenance management tasks of the testing laboratory.

[0038] Specifically, the operation and maintenance management model Ss is a full-factor digital twin model The purpose is to realize the service set required for the operation and maintenance management tasks of the testing laboratory. It includes the technical service module TS, the user service module UT and the functional application module FA. Among them, the operation and maintenance management model is TS is the technical service module, UT is the user service module, and FA is the functional application module. The technical service module is a collection of technical services required for functional operation. The technical service module includes the model service module for model building, rendering, baking and other model services. ;Data service modules for data services such as database design, data analysis, processing and storage ; Comprehensive service module for comprehensive services such as interface encapsulation, communication protocol, algorithm verification, etc. User service module It is the service support required by the LIMS system built around the full-factor resources of man-machine-material-method-environment-test-audit in the whole process of testing. Figure 4 As shown, it includes customer entrustment - task registration and issuance - sample delivery - test scheduling and execution - report preparation / review / issuance / archiving. The whole process of testing includes testing tasks, testing equipment management, testing sample management, testing process monitoring, testing data optimization, testing environment configuration, operation and maintenance decision analysis, online report generation and operation and maintenance guidance training. Functional application module FA is used for the functional services required for the operation and maintenance management process of the testing laboratory, including laboratory operation, laboratory management, laboratory equipment management and laboratory data analysis.

[0039] Step 5: Construct a learning decision model to analyze the status of the testing equipment in the testing laboratory based on the real-time data received by the data interaction model.

[0040] Specifically, the learning decision model is , It is the execution control based on the action strategy selected according to the detection task and laboratory environment information. The reward and punishment mechanism includes reward function and penalty function; Set conditions for the algorithm; Execution control based on action strategies selected according to the detection task and laboratory environment information Includes: installation of samples to be tested, parameter setting, performance testing, environmental adaptability testing, data collection and analysis, and report preparation.

[0041] Reward and Punishment Mechanism To execute control Evaluation mechanism, reward and punishment mechanism The reward function in is used to motivate the operator to select the current action strategy signal, and the penalty function is used to adjust the operator's current action strategy signal. The reward function includes action reward, position reward and distance reward, and the penalty function includes collision penalty, in-area penalty and time penalty.

[0042] Algorithm Setting Conditions The conditions that need to be set to complete learning and decision-making training include algorithm network structure design, training parameter setting, and round termination condition setting.

[0043] The learning decision model LD is a full-factor digital twin model Provide data-driven methods and intelligent algorithms to solve hidden faults in the testing equipment of the testing laboratory. Usually, hidden faults mostly occur in the operating mechanism composed of the mechanical system and electrical system inside the testing equipment. Fault factors include bearing vibration, motor voltage and current, motor resistance temperature, etc. The fault factors are complex and there are coupling effects. It is difficult for the operation and maintenance managers of the testing laboratory to perceive and judge only by the appearance of the components.

[0044] The construction process of learning decision model LD is as follows Figure 5 As shown in the figure, through the real-time collection of multi-source data such as the operating status of the equipment in the testing laboratory, equipment detection tasks, and the fault status of key equipment and components, training sample data is constructed with the historical detection equipment data in the database, and feature extraction algorithms such as stack denoising autoencoder and support vector machine are used to obtain the relevant fault characteristics of the detection equipment and determine the key coupling factors, and then build random forest, GBDT, XGBOOST and other algorithms based on gradient ascent for model training to achieve predictive maintenance of detection equipment.

[0045] After receiving the real-time data transmitted by the data interaction model, the learning decision model LD matches the real-time data from different sources with the algorithms in the preset algorithm set, and analyzes the detection equipment in the detection laboratory for fault diagnosis, anomaly detection, remaining service life prediction, and multi-objective optimization based on the matched algorithms. For the hidden faults that are difficult to detect and judge in the detection equipment, the machine learning algorithm is used to establish an intelligent operation and maintenance decision algorithm set, match the multi-source heterogeneous real-time data with the algorithm set, and continuously iterate and optimize through online learning decision and offline algorithm training and associate it with the operation and maintenance management model Ss in the system, thereby reducing the difficulty of operation and maintenance.

[0046] Step 6: Based on the constructed physical object model of the testing laboratory, the virtual model of the testing laboratory, the operation and maintenance management model, the learning decision model and the full-factor data model, analyze the data interaction relationship between the various models of the testing laboratory and build a data interaction model of the testing laboratory.

[0047] Specifically, the data interaction model CN ​​is a full-factor digital twin model The artery is the core of realizing the dynamic operation of the system and the integration of virtual and real space. Among them, the data interaction model is: For the interaction between PM and DD, It is to combine multiple sensor nodes to form a self-organizing network to collect information of the physical object model PM of the detection laboratory in real time, and build a low-cost communication network based on industrial Ethernet for the detection laboratory equipment layer (equipment controller), collection layer (data acquisition hardware / system), and management layer (data acquisition management platform), and transmit the collected data of the physical object model PM of the detection laboratory to the full-factor data model DD. At the same time, For the interaction between PM and Ss, each controller in the physical object model PM of the detection laboratory can receive feedback data and generate control instructions to adjust and optimize the operation and maintenance management process of the detection laboratory.

[0048] For the interaction between PM and VM, Use and In a similar implementation, the real-time collected data is transmitted to the virtual model VM of the detection laboratory, and the virtual model VM of the detection laboratory is driven to perform dynamic simulation. At the same time, the simulation results, predictions, forecasts, and decision-making plans of the virtual model of the detection laboratory are converted into relevant control instructions and transmitted to the corresponding physical entities in the physical object model PM of the detection laboratory for real-time control and management.

[0049] For the interaction between VM and DD, It establishes a connection with the database through interface technologies such as JDBC and ODBC, and transmits data such as learning strategies, simulations, and management plans to the full-factor data model DD for storage in real time. It can also read the latest data in the database to drive the construction of the virtual model VM of the detection laboratory.

[0050] is the interaction between VM and Ss, The data connection between the detection laboratory virtual model VM and the operation and maintenance management model Ss is established through communication mechanisms such as TCP / IP, so as to realize data transmission and reception, synchronous update, and instruction transmission between the detection laboratory virtual model VM and the operation and maintenance management model Ss.

[0051] is the interaction between Ss and DD, Use and Similar technologies establish real-time connections with databases, transmit data generated during the operation and maintenance management model Ss operation to the database for storage, and can also read historical data, model parameters, and commonly used algorithms in the database in real time to support the operation and optimization of the operation and maintenance management model Ss.

[0052] is the interaction between DD and LD, It uses communication mechanisms such as Socket to connect and interact with the learning environment and the external trainer, combines twin data and environmental information for learning and decision-making to obtain the optimal management strategy, and stores the learning results in the twin database.

[0053] Step 7: Based on the constructed models of the testing laboratory, analyze the correlation between the testing process for the entire testing process and all testing elements, and build an element digital twin model to enable staff to build an intelligent operation and maintenance system for the testing laboratory based on the full-element digital twin model.

[0054] Specifically, the full-factor digital twin model is .

[0055] The full-factor digital twin model provided by the embodiment of the present invention includes a multi-level twin model of the digital twin model of the detection equipment, the digital twin model of the detection process, and the digital twin model of the detection laboratory operation and maintenance management.

[0056] The embodiment of the present invention constructs a full-factor digital twin model of the detection laboratory operation and maintenance system, and uses sensors, PLC and other data acquisition equipment to build a detection laboratory virtual model that can truly reflect the real-time status of the laboratory's physical object model, realize the mapping from the detection laboratory physical object model to the detection laboratory virtual model, and collect the data sets generated during the test process in real time, further improving the detection laboratory operation and maintenance efficiency.

[0057] The embodiment of the present invention designs a testing laboratory operation and maintenance system based on digital twins based on the above-mentioned full-factor digital twin model. Figure 6 As shown in Figure 1, the system architecture mainly includes: testing laboratory, system support layer, data collection layer and application display layer. Figure 7 Design according to the system construction process shown in the figure.

[0058] Combined with the functional requirements of the testing laboratory, the functional modules of the testing laboratory operation and maintenance system include laboratory operation module, laboratory management module, test equipment management module and test data analysis module. Figure 8As shown, the testing laboratory operation and maintenance system includes 4 major functional modules and 23 functions, which can realize data collection, data monitoring, data query, data curve, data statistics, data report, system integration and general functions of the testing laboratory, and support more than 50 standard communication protocols of mainstream control systems. Based on the overall system architecture and related functional modules, the test laboratory operation and maintenance system is developed and deployed in a front-end and back-end separation manner. The front-end is developed based on technologies such as React, zustand, and less to receive and display equipment data, and the back-end provides equipment data interfaces and real-time mqtt data. The operation and maintenance system can realize the operation and maintenance management of the test laboratory, test equipment management, and test data analysis.

[0059] Based on digital twin technology, the present invention develops an intelligent operation and maintenance system for testing laboratories based on digital twins through means of virtual-reality interactive feedback, data fusion analysis, decision-making iterative optimization and other means in the full-factor digital twin model. It monitors, diagnoses, simulates, makes decisions, and controls the entire life cycle of the physical entities in the testing laboratory, realizes visual operation and maintenance management of the testing process and digital transformation and upgrading of the testing laboratory.

[0060] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0061] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0062] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for constructing a full-factor digital twin model of a testing laboratory, characterized in that: include: Step 1: Analyze the physical objects and data collection objects of the detection laboratory and build a physical object model of the detection laboratory. EO is the physical object of the testing laboratory, and DCO is the data collection object of the testing laboratory; the physical object of the testing laboratory includes the laboratory layout, testing equipment, tested samples, testing personnel, and testing conditions; the data collection object of the testing laboratory includes data collection equipment and data transceiver modules. The data collection equipment includes laboratory data collection equipment for testing related sensors and PLCs, and the data transceiver modules include data collection gateways and industrial switches; Step 2: Based on the physical object model of the test laboratory, a virtual model of the test laboratory based on digital twin is constructed. GM is the geometric model of the testing laboratory; RM is the rule model of the testing laboratory; FM is the process model of the testing laboratory; the laboratory geometric model of the test environment and testing equipment is constructed by means of 3D scanning modeling, point cloud data processing and mesh surface rendering; the laboratory process model is established by combing the mechanism coupling relationship between the basic information of the testing laboratory and the testing process information; the laboratory rule model is constructed through the laboratory geometric model, laboratory process model and the operation and maintenance system functions of multi-equipment mechanism linkage and cross-regional test data sharing; Step 3: Based on the test laboratory data sets of each model, construct the full-factor data model of the test laboratory, where the full-factor data model is ; a test lab environment configuration data set generated when running the test lab physical object model; The basic data of the test equipment, the simulation data of the geometric model and the test result data of the process model when running the virtual model of the test laboratory; and They are respectively the operation and maintenance status data and operation and maintenance knowledge data generated when running the operation and maintenance management system and function display; In order to run the learning decision model, the above multi-source data is autonomously analyzed, learned and decided to form a decision optimization data set; Step 4: According to the operation and maintenance management tasks of the testing laboratory, build an operation and maintenance management model, where the operation and maintenance management model is , TS is the technical service module, UT is the user service module, and FA is the functional application module; The technical service module is a collection of technical services required for the function operation. The technical service modules are: For the model service module, For the data service module, It is a comprehensive service module; The user service module is a service support built around the full-factor resources of man-machine-material-method-environment-test-audit in the whole process of testing; the functional application module is used to integrate the functional services required in the operation and maintenance management of the testing laboratory; Step 5: Construct a learning decision model to analyze the status of the testing equipment in the testing laboratory based on the real-time data received by the data interaction model. The learning decision model is , It is the execution control based on the action strategy selected according to the detection task and laboratory environment information. The reward and punishment mechanism includes reward function and penalty function. Set conditions for the algorithm; Step 6: Based on the constructed physical object model of the testing laboratory, the virtual model of the testing laboratory, the operation and maintenance management model, the learning decision model and the full-factor data model, the data interaction relationship between the various models of the testing laboratory is analyzed, and the data interaction model of the testing laboratory is constructed. , , and They are the real-time data interaction between the physical object model of the testing laboratory and the full-factor data model, the virtual model of the testing laboratory and the operation and maintenance management model; and They are the real-time data interaction between the virtual model of the testing laboratory and the full-factor data model and the operation and maintenance management model respectively; To provide real-time data interaction between the operation and maintenance management model and the full-factor data model; Real-time data interaction between the full-factor data model and the learning decision model; Step 7: Based on the constructed models of the testing laboratory, analyze the correlation between the testing process and all the elements of the testing process, and build a digital twin model of the elements, so that the staff can build an intelligent operation and maintenance system for the testing laboratory based on the full-element digital twin model. The full-element digital twin model is .

2. The method for constructing a full-factor digital twin model of a detection laboratory according to claim 1, characterized in that: In step 2, the process model is used to control the docking process of the overall business process of the testing laboratory; The rule model is used to describe and characterize the test information and detection process of the sample to be tested; The basic information for constructing the process model includes: laboratory environment, test conditions, testing equipment and operators; Testing process information includes test commissioning, test execution and report release.

3. The method for constructing a full-factor digital twin model of a detection laboratory according to claim 1, characterized in that: In step 3, the test laboratory environment configuration data set includes: test equipment operation status, test personnel configuration, test cycle, laboratory power supply, ventilation and temperature and humidity environment; The operation and maintenance status data generated during the operation of the operation and maintenance management system includes data type conversion, preprocessing, classification and integration; Operation and maintenance knowledge data includes inspection records, repair and maintenance records, knowledge and experience of testing experts, testing equipment manuals and laboratory management plans.

4. The method for constructing a full-factor digital twin model of a testing laboratory according to claim 1 is characterized in that: The process of the learning decision model analyzing the status of the testing equipment in the testing laboratory based on the real-time data received by the data interaction model includes: After the learning decision model receives the real-time data transmitted by the data interaction model, it matches the real-time data from different sources with the algorithms in the preset algorithm set, and performs fault diagnosis, anomaly detection, remaining service life prediction and multi-objective optimization on the testing equipment in the testing laboratory based on the matched algorithms.

5. The method for constructing a full-factor digital twin model of a testing laboratory according to claim 1 is characterized in that: In step 5, the execution control is performed based on the action strategy selected according to the detection task and the laboratory environment information, including: installation of the sample to be tested, parameter setting, performance testing, environmental adaptability testing, data collection and analysis, and report preparation; The reward function in the reward and punishment mechanism is used to motivate the operator to select the current action strategy signal, and the penalty function is used to adjust the operator's current action strategy signal. The reward function includes action reward, position reward and distance reward, and the penalty function includes collision penalty, intra-region penalty and time penalty. The algorithm setting conditions are the conditions that need to be set to complete learning and decision-making training, including algorithm network structure design, training parameter setting, and round termination condition setting.

6. The method for constructing a full-factor digital twin model of a detection laboratory according to any one of claims 1 to 5, characterized in that: In step seven, all the testing elements include testing personnel-testing equipment-testing samples-testing methods-testing environment-testing results-report review; the entire testing process includes test environment configuration, testing personnel operation, laboratory operation and maintenance management, system intelligent decision-making and test report generation.

7. An intelligent operation and maintenance system for a testing laboratory, characterized in that: The full-factor digital twin model constructed based on the construction method of any one of claims 1 to 6 is constructed, and the system architecture includes: a testing laboratory, a system support layer, a twin model layer, and an application display layer; The system functional modules include laboratory operation module, laboratory management module, test equipment management module and test data analysis module.

Images