Industrial big data simulation verification system based on digital twinning

Abstract

The invention relates to the technical field of industrial automation and intelligent manufacturing, and discloses an industrial big data simulation verification system based on digital twinning, and the system comprises a physical entity module which is used for deploying an industrial sensor array and collecting the real-time parameter data of production equipment; the digital twin mirror image module is linked with the physical entity module through a data mapping engine, and synchronously updates a virtual mirror image model at a millisecond level according to the parameter change of the physical entity; and the data transmission module is used for transmitting the data acquired by the physical entity module to the digital twin mirror image module and the multi-source data fusion module through a 5G or industrial Ethernet, and is used for cleaning the acquired multi-modal industrial data and eliminating noise data. Through the physical entity-digital twin mirror image real-time linkage design, the limitation that traditional simulation is disjointed with entities is broken through. The industrial sensor array is combined with the edge calculation module, so that rigid linkage of physical production equipment, a real-time database and a digital twin mirror image is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial automation and intelligent manufacturing, and particularly to an industrial big data simulation verification system based on digital twinning. BACKGROUND

[0002] In the field of industrial production, with the acceleration of digital transformation, the simulation verification technology of industrial big data has become a key means to improve production efficiency, optimize product quality and reduce production cost. Traditional industrial simulation technology mainly relies on physical models and computer simulation, and predicts and verifies various working conditions in the production process by establishing virtual models. These technologies have improved the controllability and predictability of the production process to some extent, but as the complexity of industrial production and the amount of data increase, the existing technology gradually exposes some limitations. First, there is a delay in the data synchronization between physical devices and virtual models in traditional simulation technology, which leads to deviations between simulation results and actual production conditions, affecting the accuracy of production decisions. Second, the data generated in the industrial production process has the characteristics of multi-modal and high noise, and the existing technology often has difficulty in effectively cleaning and fusing these complex data, resulting in inaccurate calibration of simulation parameters. In addition, the existing technology relies on a large number of physical tests in complex working condition verification, which not only increases the cost of testing, but also consumes a large amount of energy, and lacks intelligent scheduling mechanism in the allocation of computing resources, resulting in low resource utilization. Finally, after the simulation verification is completed, the existing technology usually requires manual intervention for result analysis and optimization scheme formulation, increasing the workload and possibly delaying the implementation of the optimization scheme, reducing the intelligent level of the production process.

[0003] These problems limit the further development and application of simulation verification technology in industrial production, therefore, a new simulation verification system is needed to solve the above problems to meet the needs of modern industrial production for high precision, high efficiency and low energy consumption. SUMMARY

[0004] In view of the deficiencies of the prior art, the present application provides an industrial big data simulation verification system based on digital twinning, which solves the problems of asynchronization between physical devices and virtual models, low verification accuracy, low resource utilization and lack of full-process automation in traditional industrial simulation.

[0005] To achieve the above purpose, the present application realizes the following technical scheme: an industrial big data simulation verification system based on digital twinning, comprising:

[0006] a physical entity module for deploying an industrial sensor array to collect real-time parameter data of production equipment;

[0007] The digital twin mirror image module is linked with the physical entity module through a data mapping engine, and a virtual mirror image model is updated millisecond by millisecond according to parameter changes of the physical entity;

[0008] The data transmission module is configured to transmit data collected by the physical entity module to the digital twin mirror image module through 5G or industrial Ethernet;

[0009] The multi-source data fusion module is configured to clean and process the collected multi-modal industrial data, and eliminate noise data, and synchronously adjust the twin model parameter library through a data bus.

[0010] The simulation verification module is internally provided with multi-dimensional verification indexes, and can real-time feedback of simulation results and physical data deviation, drive model parameter gradient correction, and ensure that the simulation verification error is within a preset range.

[0011] The resource optimization module includes a simulation scene scheduler, a resource consumption sensor and a physical test replacement module, and is configured to reasonably allocate computing resources according to verification requirements, reduce the number of physical tests, and reduce test cost and energy consumption.

[0012] The decision optimization module includes a simulation result analysis unit, an optimization scheme generator and a physical device control interface, and is configured to generate an optimization scheme according to the simulation verification result and push it to an industrial control system, so as to realize full-process automation from simulation verification to production optimization.

[0013] Through the above technical solutions, by integrating the physical entity module, the digital twin mirror image module, the data transmission module, the multi-source data fusion module, the simulation verification module, the resource optimization module and the decision optimization module, high-precision simulation verification and full-process automation optimization of the industrial production process are realized.

[0014] Preferably, the industrial sensor array includes ten-thousand-node sensors, and the sampling frequency is 100Hz-1kHz, which is used to real-time collect parameters such as temperature, pressure and vibration of the equipment.

[0015] Through the above technical solutions, by deploying an industrial sensor array including ten-thousand-node sensors, key parameters such as temperature, pressure and vibration of the equipment are collected in real time at a sampling frequency of 100Hz-1kHz, high-precision and high-frequency monitoring of the running state of the production equipment is realized, and accurate and real-time data basis is provided for subsequent digital twin modeling, simulation verification and production optimization.

[0016] Preferably, the data cleaning engine in the multi-source data fusion module supports structured and unstructured data processing, and when the data noise fluctuation exceeds 5%, the twin model parameter library automatically calls historical calibration data for compensation.

[0017] By the technical scheme, the collected structured and unstructured data are processed by the data cleaning engine, noise data are eliminated, and when data noise fluctuation exceeds 5%, historical calibration data are automatically called for compensation, so that accurate processing and calibration of multi-source industrial data are realized, thereby improving the parameter accuracy of the digital twin model and the accuracy of simulation verification.

[0018] Preferably, the multi-dimensional verification indicators of the simulation verification module include error rate, response speed and stability.

[0019] By the technical scheme, by setting multi-dimensional verification indicators such as error rate, response speed and stability, comprehensive monitoring and accurate evaluation of the deviation between simulation results and physical data are realized, thereby ensuring high precision and reliability of simulation verification and providing strong support for optimization decision of industrial production.

[0020] Preferably, in the resource optimization module, when the CPU utilization rate exceeds 80%, the simulation scene scheduler automatically schedules non-real-time scenes to edge nodes, and the physical test replacement module performs 100% virtual verification on high-risk scenes and 1:5 ratio physical verification on low-risk scenes.

[0021] By the technical scheme, by CPU utilization monitoring, when it exceeds 80%, non-real-time simulation scenes are automatically scheduled to edge nodes, high-risk scenes are 100% virtually verified, and low-risk scenes are 1:5 ratio physically verified, so that reasonable allocation and optimized utilization of computing resources are realized, test cost and energy consumption are reduced, and simulation verification efficiency is improved.

[0022] Preferably, in the decision optimization module, the key indicators output by the simulation result analysis unit include equipment life prediction and fault risk value, and the control parameters generated by the optimization scheme generator include equipment running speed and load distribution.

[0023] By the technical scheme, by the simulation result analysis unit outputting key indicators such as equipment life prediction and fault risk value, and the optimization scheme generator generating control parameters such as equipment running speed and load distribution, full-process automation from simulation verification to production optimization is realized, and production efficiency and equipment reliability are improved.

[0024] An industrial big data simulation verification method based on digital twinning includes the following steps:

[0025] Step 1: Physical data acquisition and twin model initialization linkage adjustment, deploy an industrial sensor array to the target production equipment, the sensor is linked with the real-time database, the database stores the original data according to the preset dimension, the digital twin platform automatically generates a high-precision virtual mirror by calling the basic model library, and completes the virtual-real mapping initialization.

[0026] Step 2: Multi-source data fusion and simulation parameter calibration linkage control, the data cleaning engine fuses and processes the collected data, eliminates noise data, and transmits the signal to the twin model parameter library, the algorithm in the library automatically matches the simulation parameters, drives the parameter library to correct the material property parameters, and ensures the initial simulation accuracy;

[0027] Step 3: Multi-scenario simulation verification and resource optimization linkage execution, the simulation scenario scheduler starts multi-scenario parallel simulation according to the verification requirements, the resource consumption sensor monitors the computing load in real time, the physical test replacement module performs virtual verification on high-risk scenarios, and low-risk scenarios are selected for physical verification in proportion, the verification data is fed back to the model calibration parameter library, and the simulation accuracy is continuously optimized;

[0028] Step 4: Linkage operation of result analysis and decision optimization, after the simulation verification is completed, the analysis unit generates a comprehensive report, if the key indicators meet the standards, the optimization scheme generator automatically outputs the control parameters, the scheme is audited, the control interface links the industrial control system to issue instructions, the physical equipment executes parameter updating, the digital twin mirror synchronously records the optimization effect, and forms a closed loop.

[0029] Through the above technical scheme, by sequentially executing the steps of physical data collection and twin model initialization, multi-source data fusion and simulation parameter calibration, multi-scenario simulation verification and resource optimization, result analysis and decision optimization, full-process automation from data collection to production optimization is realized, and the efficiency, accuracy and intelligent level of industrial production are improved.

[0030] Preferably, in step 2, the data cleaning engine eliminates noise data with a deviation of >3σ, and if the vibration data fluctuates greatly, the damping coefficient calibration frequency is dynamically increased.

[0031] Through the above technical scheme, by eliminating noise data with a deviation of >3σ through the data cleaning engine, and dynamically increasing the damping coefficient calibration frequency when the vibration data fluctuates greatly, accurate processing of the collected data and real-time optimization of the simulation parameters are realized, thereby improving the accuracy and reliability of the simulation verification.

[0032] Preferably, in step 4, the key indicators include fault prediction accuracy, when the fault prediction accuracy is ≥90%, the optimization scheme generator automatically outputs the control parameters, and adjusts the equipment running speed to 1100-1200r / min.

[0033] Through the above technical scheme, by setting the fault prediction accuracy as a key indicator, device operation parameter optimization based on high-precision fault prediction is realized, and the device operation efficiency and reliability are improved.

[0034] The present application provides an industrial big data simulation verification system based on digital twinning. It has the following advantages:

[0035] 1、The application breaks through the limitation of traditional simulation disconnection with physical entity through the real-time linkage design of "physical entity-digital twin mirror image". The combination of industrial sensor array and edge computing module realizes the rigid linkage of physical production equipment, real-time database and digital twin mirror image. When the physical parameter changes, the mirror image model can be updated synchronously in milliseconds, ensuring the non-delay connection of "physical operation-virtual mapping-dynamic simulation", effectively solving the problem of traditional simulation lag. At the same time, the dynamic linkage mechanism of multi-source industrial data fusion and simulation parameters can accurately improve the verification accuracy. This high-precision simulation verification capability enables industrial enterprises to obtain more accurate data support in product design, production optimization and fault prediction, thereby improving production efficiency, reducing cost and improving product quality.

[0036] 2、The application innovatively designs a linkage mechanism of "simulation scene scheduler-resource consumption sensor-physical test replacement module", realizing intelligent linkage of simulation verification and resource optimization. In complex working condition verification, digital twin simulation is started preferentially to reduce the number of physical tests and significantly reduce test cost and energy consumption. In addition, the linkage structure of "simulation result analysis unit-optimization scheme generator-physical device control interface" at the end of the system realizes the full-process automation from simulation verification to production optimization. After simulation verification is completed, the analysis unit outputs key indicators to trigger the scheme generator to generate an optimization scheme, and the optimization parameters are pushed to the industrial control system through the control interface to realize parameter updating of the physical device. The application of this full-process automation not only improves the intelligent level of the production process, but also reduces manual intervention, improves production efficiency and system reliability. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 is the overall flowchart of the application;

[0038] Figure 2 is the simulation verification method flowchart of the application. DETAILED DESCRIPTION

[0039] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the application specification. Obviously, the described embodiments are only part of the embodiments of the application, not all. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0040] Please refer to the drawings in the application specification Figure 1 - the drawings in the application specification Figure 2 The embodiment of the application provides an industrial big data simulation verification system based on digital twin, which comprises:

[0041] A physical entity module is used to deploy an industrial sensor array to collect real-time parameter data of production equipment.

[0042] The industrial sensor array includes ten-thousand-level node sensors with a sampling frequency of 100Hz-1kHz, which are used to collect real-time parameters such as temperature, pressure, and vibration of the equipment.

[0043] A digital twin mirror module is linked with the physical entity module through a data mapping engine, and the virtual mirror model is updated millisecond-level synchronously according to the parameter changes of the physical entity.

[0044] Specifically, the physical entity module is the foundation of the entire system, which realizes real-time monitoring of production equipment by deploying an industrial sensor array. These sensor arrays contain a large number of sensor nodes, up to ten thousand, which can collect key parameters of production equipment in real time at a high sampling frequency of 100Hz to 1kHz, such as temperature, pressure, and vibration. These parameters are crucial for monitoring the running state of the equipment, as they directly reflect the performance and health status of the equipment during production.

[0045] The digital twin mirror module is closely connected with the physical entity module, and is linked through a data mapping engine. This means that when the parameters collected by the physical entity module change, the digital twin mirror module can respond quickly to update the virtual mirror model with millisecond-level high precision. This real-time updating capability ensures that the virtual model can accurately reflect the current state of the physical entity, providing a highly accurate data basis for subsequent simulation verification. In this way, the system can accurately simulate the operation of physical equipment in a virtual environment, providing reliable verification and optimization support for various complex working conditions in industrial production.

[0046] A data transmission module is used to transmit the data collected by the physical entity module to the digital twin mirror module through 5G or industrial Ethernet.

[0047] A multi-source data fusion module is used to clean and process the collected multi-modal industrial data, eliminate noise data, and synchronize adjustment with the twin model parameter library through the data bus.

[0048] The data cleaning engine in the multi-source data fusion module supports structured and unstructured data processing. When the data noise fluctuation exceeds 5%, the twin model parameter library automatically calls historical calibration data for compensation.

[0049] Specifically, the data transmission module ensures that the real-time data collected by the physical entity module can be efficiently and stably transmitted to the digital twin mirror image module. This process is achieved through 5G or industrial Ethernet, both of which have high bandwidth and low latency, ensuring the real-time and accuracy of data transmission. At the same time, the multi-source data fusion module is responsible for processing industrial data from different sources and formats. These data often exhibit multi-modal characteristics, including structured data (such as table data in databases) and unstructured data (such as text, images, etc.). The data cleaning engine plays a key role in this process, as it can effectively clean and process the collected data, eliminating noise data and improving data quality and usability. When data noise fluctuation exceeds 5%, the twin model parameter library will automatically intervene, calling historical calibration data for compensation. This mechanism ensures that even in the case of fluctuations in data quality, the system can maintain the stability and accuracy of the parameters through historical data calibration, providing a solid data foundation for subsequent simulation verification, enabling the entire system to maintain efficient and accurate operation in complex industrial environments.

[0050] The simulation verification module has multi-dimensional verification indicators, real-time feedback of simulation results and physical data deviation, driving model parameter gradient correction, ensuring that the simulation verification error is within the preset range. The multi-dimensional verification indicators of the simulation verification module include error rate, response speed, and stability.

[0051] The resource optimization module includes a simulation scene scheduler, a resource consumption sensor, and a physical test replacement module, which is used to reasonably allocate computing resources according to verification requirements, reduce the number of physical tests, and reduce test costs and energy consumption. In the resource optimization module, when the CPU utilization rate exceeds 80%, the simulation scene scheduler automatically schedules non-real-time scenes to edge nodes, the physical test replacement module performs 100% virtual verification on high-risk scenes, and low-risk scenes are physically verified at a ratio of 1:5.

[0052] Specifically, the simulation verification module and the resource optimization module are key components for efficient and accurate industrial simulation verification. The simulation verification module can monitor and feedback the deviation between simulation results and physical data in real time through built-in multi-dimensional verification indicators such as error rate, response speed, and stability. This real-time feedback mechanism enables the system to quickly identify differences between simulation models and actual production, and adjust the simulation model through gradient correction of model parameters, ensuring that the simulation verification error remains within the preset range. This dynamic adjustment capability greatly improves the accuracy and reliability of simulation verification, providing more accurate data support for decision-making in industrial production.

[0053] Meanwhile, the resource optimization module is dedicated to improving the resource utilization efficiency of the entire system. This module includes a simulation scenario scheduler, a resource consumption sensor, and a physical test substitution module. The simulation scenario scheduler can reasonably allocate computing resources based on the current verification requirements and the use of system resources. When the CPU utilization rate exceeds 80%, the scheduler will automatically schedule non-real-time simulation scenarios to the edge node, thereby avoiding the excessive concentration and waste of computing resources. In addition, the physical test substitution module conducts 100% virtual verification in high-risk scenarios, and conducts physical verification sampling according to a 1:5 ratio for low-risk scenarios. This strategy not only reduces the number of physical tests, reduces test costs and energy consumption, but also improves the efficiency and safety of the verification process. Through these innovative designs, the invention significantly improves resource utilization efficiency while ensuring simulation verification accuracy, bringing significant economic and environmental benefits to industrial production.

[0054] The decision optimization module includes a simulation result analysis unit, an optimization scheme generator, and a physical device control interface, which is used to generate optimization schemes based on simulation verification results and push them to the industrial control system, realizing the full-process automation from simulation verification to production optimization. In the decision optimization module, the key indicators output by the simulation result analysis unit include device life prediction and fault risk value; the control parameters generated by the optimization scheme generator include device running speed, load distribution, etc.

[0055] Specifically, the decision optimization module is composed of a simulation result analysis unit, an optimization scheme generator, and a physical device control interface, which work together to ensure the intelligence and efficiency of the production process. The simulation result analysis unit is responsible for in-depth analysis of the data after simulation verification, and outputs key indicators including device life prediction and fault risk value. These indicators provide important data support for production decisions, helping enterprises to plan equipment maintenance and production scheduling in advance, thereby reducing the risk of production interruption caused by equipment failure. The optimization scheme generator generates specific control parameters such as device running speed and load distribution based on the key indicators provided by the analysis unit. The adjustment of these parameters aims to optimize the production process, improve production efficiency, and reduce energy consumption and costs. Finally, the physical device control interface pushes the generated optimization scheme to the industrial control system in real time, ensuring that physical devices can be adjusted and operated according to the latest optimization parameters. Through this series of automated processes, the invention not only improves the intelligence level of the production process, but also realizes the real-time and dynamic nature of production optimization, bringing significant economic benefits and production efficiency improvements to enterprises.

[0056] An industrial big data simulation verification method based on digital twinning, comprising the following steps:

[0057] Step 1: Physical data acquisition and twin model initialization linkage adjustment, deploy industrial sensor array to target production equipment, sensor linkage with real-time database, database stores raw data according to preset dimensions, digital twin platform calls basic model library to automatically generate high-precision virtual mirror, complete virtual-real mapping initialization;

[0058] Step 2: Multi-source data fusion and simulation parameter calibration linkage control, data cleaning engine performs fusion processing on collected data, removes noise data, signal transmission to twin model parameter library, library algorithm automatically matches simulation parameters, drives parameter library to correct material property parameters, ensures initial simulation accuracy; data cleaning engine removes noise data with deviation > 3σ value, if vibration data fluctuation is large, dynamically improve damping coefficient calibration frequency.

[0059] Step 3: Multi-scenario simulation verification and resource optimization linkage execution, simulation scenario scheduler starts multi-scenario parallel simulation according to verification requirements, resource consumption sensor real-time monitors computing load, physical test replacement module performs virtual verification on high-risk scenarios, low-risk scenarios are proportionally selected for physical verification, verification data is fed back to model calibration parameter library, continuously optimizes simulation accuracy;

[0060] Step 4: Linkage operation of result analysis and decision optimization, after simulation verification is completed, analysis unit generates comprehensive report, if key indicators meet the standard, optimization scheme generator automatically outputs control parameters, after the scheme is audited, control interface links industrial control system to issue instructions, physical equipment executes parameter update, digital twin mirror synchronously records optimization effect, forms a closed loop. Key indicators include fault prediction accuracy, when fault prediction accuracy ≥ 90%, optimization scheme generator automatically outputs control parameters, adjusts equipment running speed to 1100-1200 r / min.

[0061] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, the scope of the present application being defined by the appended claims and their equivalents.

Claims

1. The industrial big data simulation verification system based on digital twins is characterized by: include: The physical entity module is used to deploy industrial sensor arrays to collect real-time parameter data of production equipment; The digital twin mirror module is linked with the physical entity module through a data mapping engine to synchronously update the virtual mirror model in milliseconds according to the parameter changes of the physical entity; The data transmission module is used to transmit the data collected by the physical entity module to the digital twin mirror module via 5G or industrial Ethernet; The multi-source data fusion module is used to clean and process the collected multi-modal industrial data, remove noise data, and synchronize the data with the twin model parameter library through the data bus. The simulation verification module has built-in multi-dimensional verification indicators, provides real-time feedback on the deviation between simulation results and physical data, drives the gradient correction of model parameters, and ensures that the simulation verification error is within the preset range; Resource optimization module, including simulation scenario scheduler, resource consumption sensor and physical test replacement module, is used to reasonably allocate computing resources according to verification requirements, reduce the number of physical tests, and reduce test costs and energy consumption; The decision optimization module, including a simulation result analysis unit, an optimization solution generator, and a physical equipment control interface, is used to generate optimization solutions based on simulation verification results and push them to the industrial control system, realizing full process automation from simulation verification to production optimization.

2. The industrial big data simulation verification system based on digital twins according to claim 1 is characterized in that: The industrial sensor array includes 10,000-level node sensors with a sampling frequency of 100Hz-1kHz, which is used to collect parameters such as temperature, pressure, vibration, etc. of the equipment in real time.

3. The industrial big data simulation verification system based on digital twin according to claim 1 is characterized in that: The data cleaning engine in the multi-source data fusion module supports structured and unstructured data processing. When the data noise fluctuation exceeds 5%, the twin model parameter library automatically calls historical calibration data for compensation.

4. The industrial big data simulation verification system based on digital twin according to claim 1 is characterized in that: The multi-dimensional verification indicators of the simulation verification module include error rate, response speed, and stability.

5. The industrial big data simulation verification system based on digital twins according to claim 1 is characterized in that: In the resource optimization module, when the CPU utilization rate exceeds 80%, the simulation scenario scheduler automatically schedules non-real-time scenarios to edge nodes. The physical test replacement module performs 100% virtual verification on high-risk scenarios and selects physical verification for low-risk scenarios at a ratio of 1:

5.

6. The industrial big data simulation verification system based on digital twin according to claim 1 is characterized in that: In the decision optimization module, the key indicators output by the simulation result analysis unit include equipment life prediction and failure risk value; the control parameters generated by the optimization solution generator include equipment operating speed, load distribution, etc.

7. A digital twin-based industrial big data simulation verification method, according to the digital twin-based industrial big data simulation verification system according to any one of claims 1 to 6, characterized in that: The following steps are involved: Step 1: Physical data acquisition and twin model initialization are linked and adjusted. The industrial sensor array is deployed to the target production equipment. The sensors are linked to the real-time database. The database stores the raw data according to the preset dimensions. The digital twin platform calls the basic model library to automatically generate a high-precision virtual image and complete the initialization of the virtual-real mapping. Step 2: Multi-source data fusion and simulation parameter calibration are linked and controlled. The data cleaning engine fuses the collected data, removes noise data, and transmits the signal to the twin model parameter library. The algorithm in the library automatically matches the simulation parameters and drives the parameter library to correct the material property parameters to ensure the initial simulation accuracy. Step 3: Multi-scenario simulation verification and resource optimization are executed in conjunction. The simulation scenario scheduler initiates multi-scenario parallel simulation based on verification requirements. Resource consumption sensors monitor computing load in real time. The physical test replacement module performs virtual verification on high-risk scenarios. Low-risk scenarios are proportionally sampled for physical verification. Verification data is fed back into the model calibration parameter library to continuously optimize simulation accuracy. Step 4: Linking result analysis and decision optimization. After the simulation verification is completed, the analysis unit generates a comprehensive report. If the key indicators meet the standards, the optimization solution generator automatically outputs the control parameters. After the solution is reviewed, the control interface links the industrial control system to issue instructions, the physical equipment executes the parameter update, and the digital twin mirror synchronously records the optimization effect, forming a closed loop.

8. The industrial big data simulation verification method based on digital twins according to claim 7 is characterized in that: In step 2, the data cleaning engine removes noise data with a deviation greater than 3σ. If the vibration data fluctuates greatly, the damping coefficient calibration frequency is dynamically increased.

9. The industrial big data simulation verification method based on digital twins according to claim 7 is characterized in that: In step 4, the key indicator includes the fault prediction accuracy. When the fault prediction accuracy is ≥90%, the optimization solution generator automatically outputs the control parameters and adjusts the equipment operating speed to 1100-1200 r / min.

Images