Industrial fluid mechanics multi-field coupling real-time monitoring and feedback simulation system

Abstract

The invention discloses an industrial fluid mechanics multi-field coupling real-time monitoring and feedback simulation system, and relates to the technical field of fluid mechanics monitoring and feedback simulation systems. The data acquisition module uses a self-adaptive multi-mode sensor array and a built-in chip for edge calculation; the data transmission module is based on an SDN and a block chain, and uses an error correction code to ensure safety and reliability; the processing and analysis module combines deep learning and a traditional method and optimizes parameters through reinforcement learning; the simulation module adopts VR / AR and dynamic grid technologies; and the feedback control module adopts FAMPC and genetic algorithm optimization control. According to the invention, multi-modal acquisition and high-speed safe transmission guarantee data quality, deep learning and a traditional method are combined for precise analysis, intelligent control and adaptive adjustment are realized, fault diagnosis and early warning are timely, modular design is easy to maintain, integration with other systems is facilitated, and industrial intelligent efficient production is assisted.

Description

Technical Field

[0001] This invention relates to the field of fluid dynamics monitoring and feedback simulation systems, and in particular to a real-time monitoring and feedback simulation system for multi-field coupling in industrial fluid dynamics. Background Technology

[0002] In modern industrial production, industrial fluid systems are widely used in many fields such as petrochemicals, power energy, and aerospace. Their operating status directly affects production efficiency, product quality, and production safety. The operation of industrial fluid systems involves the coupling effects of multiple physical fields, including fluid mechanics, thermodynamics, and electromagnetics. Changes in parameters such as fluid pressure, temperature, and flow velocity, as well as interference from factors such as electric and magnetic fields, all have a significant impact on system performance. Accurate monitoring and analysis of the multi-field coupling state of industrial fluid systems are of great significance for ensuring stable system operation and optimizing production processes.

[0003] Traditional industrial fluid monitoring methods primarily rely on single-type sensors for data acquisition, only obtaining partial physical parameters of the fluid and failing to comprehensively reflect the complex state of multi-field coupling. Furthermore, the acquired data often suffers from low accuracy and high noise, making it difficult to meet the demands of high-precision monitoring. In terms of data transmission, traditional communication networks suffer from low transmission rates and poor stability, unable to achieve real-time data transmission, resulting in delayed monitoring results and the inability to promptly detect anomalies during system operation.

[0004] In terms of data processing and simulation, most existing systems employ single numerical calculation methods or simple empirical models, which are insufficient to accurately describe the complex physical processes of multi-field coupling, resulting in significant deviations between simulation results and actual conditions. Feedback control is also relatively crude, typically employing fixed-parameter control strategies that cannot adaptively adjust to the system's real-time operating status, leading to poor control accuracy and stability. Furthermore, existing industrial fluid monitoring and simulation systems lack intelligent fault diagnosis and early warning functions, failing to predict potential faults in advance. Once a fault occurs, it may lead to production interruptions, equipment damage, or even safety accidents, causing substantial economic losses and social impact. Therefore, there is an urgent need to develop an industrial fluid dynamics system capable of real-time monitoring, accurate simulation, and intelligent feedback control of multi-field coupling. Summary of the Invention

[0005] This invention proposes an industrial fluid dynamics multi-field coupling real-time monitoring and feedback simulation system to solve the problems mentioned in the prior art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an industrial fluid mechanics multi-field coupling real-time monitoring and feedback simulation system, comprising:

[0007] Data acquisition module: Employs an adaptive sensor array to dynamically adjust the sensor's acquisition frequency and operating mode, automatically switching sensing modes according to different physical fields. The acquisition frequency is adjusted using a formula. Determined, where ΔL is the minimum resolvable distance of the sensor, v max ρ is the maximum flow velocity of the fluid, γ is the density change influence coefficient, Δρ is the density change, and ρ base It is based on the base density; at the same time, the sensor has a built-in microchip to perform preliminary edge computing to remove noise and interference from the collected data, and to extract features;

[0008] Data transmission module: Utilizing adaptive transmission technology based on Software-Defined Networking (SDN), it monitors bandwidth, latency, and packet loss rate parameters of each communication path in real time, and dynamically adjusts data transmission paths and bandwidth allocation based on these parameters. Simultaneously, it employs encryption algorithms to encrypt data and introduces a data error correction code mechanism during data transmission, using formulas... Calculate the optimal transmission time T optimal Data size For data size, Bandwidth i Let the bandwidth of the i-th path be Loss. i Let be the packet loss rate of the i-th path;

[0009] Data Processing and Analysis Module: This module constructs a hybrid computing architecture based on deep learning and multiphysics models. It combines Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) learning models with finite element methods and finite difference methods to process and analyze multiphysics coupled data. Reinforcement learning algorithms are introduced to optimize the parameters of the multiphysics coupled model in real time using formulas. Update the model parameters θ, where α is the learning rate, A(s, a) is the advantage function, and π θ (a|s) is the policy function.

[0010] Furthermore, it also includes:

[0011] Simulation Module: Employing real-time interactive simulation technology based on virtual reality and augmented reality, users participate in the simulation process of industrial fluid systems through the device, adjusting simulation parameters in real time. The system updates simulation results in real time and displays them to the user in a 3D visualization format. Simultaneously, physically based rendering (PBR) technology is introduced to ensure the simulation results accurately reflect reality. During the simulation, dynamic mesh adaptive technology is used to automatically adjust the mesh density and shape according to the fluid flow state, using formulas... Determine the mesh density. density Where ΔV is the rate of change of fluid volume. The magnitude of the velocity gradient;

[0012] Feedback Control Module: This module proposes an optimized control strategy based on Fuzzy Adaptive Model Predictive Control (FAMPC) to adaptively adjust control parameters according to the real-time operating status and simulation results of the industrial fluid system. Simultaneously, a genetic algorithm is introduced to optimize the FAMPC parameters using formulas. Calculate the optimal control input Where y k+i|k To predict the output, y ref，k+i For the reference output, Q and R are weight matrices;

[0013] Fault Diagnosis and Early Warning Module: This module employs a fault diagnosis method based on the fusion of Deep Belief Networks (DBN) and evidence theory, while also introducing a fault early warning mechanism based on time series analysis. By analyzing historical and real-time data, it predicts the probability and timing of fault occurrence; and uses formulas... Calculate the probability of failure P fault , where w i D represents the weight of the i-th fault feature. i For the diagnostic result of the i-th fault feature, λ i t is the attenuation coefficient of the i-th fault feature, and t is time; when the probability of fault occurrence exceeds the preset threshold, an early warning signal is issued, and a fault diagnosis report and solution suggestions are provided.

[0014] Database management module: Employs a hybrid storage architecture based on graph databases and distributed file systems, while also incorporating semantic search-based query technology. Users describe their query requirements in natural language, and the system automatically parses the query statement and obtains the results; through formulas... Calculate the score of the query results. result , where w i The weight of the i-th query keyword, Similarity i To query the similarity between keywords and results.

[0015] Furthermore, the sensor array of the data acquisition module adopts a biomimetic design concept, and the sensor array is equipped with self-healing capabilities. When a sensor malfunctions, it automatically adjusts the sensing mode and data acquisition strategy, through a formula. The self-healing capability S of the computational sensor repair S original For the original sensitivity, S fault The sensitivity after a fault is given, μ is the compensation coefficient of the redundant sensor, and N is the value of N. redundant N represents the number of redundant sensors. total This represents the total number of sensors.

[0016] Furthermore, the data transmission module employs network slicing technology based on Software-Defined Networking (SDN), dividing network resources into virtual network slices. Each slice is customized according to different service requirements and quality of service (QoS) requirements. Simultaneously, a network slice management mechanism based on distributed ledger technology is introduced, using formulas... Calculate the Quality of Service (QoS) of network slices slice , where w i The weight of the i-th quality of service metric, QoS i Let be the value of the i-th service quality indicator.

[0017] Furthermore, the data processing and analysis module adopts a hybrid programming model based on a heterogeneous computing platform, combining the advantages of different computing devices such as CPU, GPU, and FPGA to achieve the solution of multi-field coupled models and the computation of learning algorithms; it uses OpenMP, CUDA, and OpenCL parallel programming technologies to parallelize the computing tasks; and it introduces a resource management mechanism based on task scheduling algorithms to dynamically allocate computing resources according to the characteristics of the computing tasks and the performance of the computing devices; through formulas... Computational resource utilization efficiency resource Task completed To the number of computational tasks completed, Rsource used This represents the amount of computing resources used.

[0018] Furthermore, the simulation module employs a collaborative simulation technique based on multi-scale modeling and multi-physics coupling to couple the industrial fluid system model with the fluid molecule model, considering the influence of fluid structure and molecular interactions on flow characteristics. Simultaneously, it couples fluid mechanics, electromagnetics, and thermodynamics to achieve realistic simulation. During the collaborative simulation process, distributed computing technology based on the message passing interface (MPI) is used to improve simulation efficiency. This is achieved through formulas... Error in calculating simulation results simulation , where Result simulation The simulation results are shown in the Result column. experiment These are the experimental results.

[0019] Furthermore, the feedback control module adopts a distributed cooperative control strategy based on a multi-agent system (MAS). Each control unit in the industrial fluid system is abstracted as an agent, and each agent is configured with autonomous decision-making and cooperative processing capabilities. Distributed cooperative control of the industrial fluid system is achieved through information interaction and negotiation between agents. Simultaneously, a game theory-based agent decision-making mechanism is introduced, using formulas... Payoff for the intelligent agent agent , where w iLet Utility be the weight of the i-th utility index. i Let be the value of the i-th utility index.

[0020] Furthermore, it also includes:

[0021] Remote monitoring and management module: Adopting an architecture based on the integration of cloud computing and edge computing, some data processing and analysis tasks are distributed to edge nodes for processing; at the same time, storage and computing resources are transmitted to the cloud server to realize centralized management and sharing of data; users can remotely access the system through mobile terminals and web browsers to view the operating status, simulation results and control command information of the industrial fluid system in real time.

[0022] Furthermore, the system adopts a modular design approach, ensuring independence and scalability among modules. It also introduces system integration technology based on microservice architecture, encapsulating each module as an independent microservice that interacts through communication protocols. In addition, the system provides standardized interfaces and open APIs to enable integration and data sharing with industry systems.

[0023] Compared with existing technologies, the beneficial effects of this invention are:

[0024] For data acquisition, an adaptive multimodal sensing sensor array and edge computing technology are employed, which not only accurately captures information from multiple fields but also processes data in real time, significantly improving the quality and efficiency of the acquired data. The data transmission module utilizes SDN and blockchain technologies to achieve high-speed, secure, and reliable data transmission, ensuring timely and accurate delivery of information.

[0025] The data processing and analysis module combines deep learning and traditional numerical computation methods to deeply mine data features and accurately construct multi-field coupled models, effectively improving the ability to analyze and predict the operating status of industrial fluid systems. The simulation module utilizes VR / AR and advanced rendering technologies to provide an immersive and highly realistic simulation experience, while dynamic mesh adaptive technology ensures high accuracy and efficiency in the simulation.

[0026] The feedback control module employs an intelligent optimization strategy, adaptively adjusting control parameters based on the system's real-time status, significantly improving control accuracy and stability. The fault diagnosis and early warning module integrates multiple intelligent algorithms to achieve accurate fault diagnosis and early warning, reducing the risk of accidents. The database management module's hybrid storage architecture and intelligent query technology facilitate data storage, management, and retrieval. The remote monitoring and management module, based on a cloud-edge converged architecture, supports multiple remote access methods, with blockchain technology ensuring access security. The modular and microservice architecture design makes the system easy to upgrade and maintain, and facilitates integration with other systems, providing strong support for the intelligent and efficient development of industrial production. Attached Figure Description

[0027] Figure 1 This is a schematic block diagram of a real-time monitoring and feedback simulation system for multi-field coupling in industrial fluid mechanics proposed in this invention. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0030] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The invention will now be described in further detail with reference to the accompanying drawings.

[0031] Reference Figure 1 : A specific implementation of a real-time monitoring and feedback simulation system for multi-field coupling in industrial fluid dynamics

[0032] I. System Overall Architecture Overview This industrial fluid mechanics multi-field coupling real-time monitoring and feedback simulation system aims to achieve real-time, accurate monitoring and feedback control of the multi-field coupling state of industrial fluid systems. It is composed of a data acquisition module, a data transmission module, a data processing and analysis module, a simulation module, a feedback control module, a fault diagnosis and early warning module, a database management module, and a remote monitoring and management module.

[0033] II. Detailed Implementation Process of Each Module

[0034] (I) Data Acquisition Module

[0035] Sensor arrays with adaptive multimodal sensing capabilities are deployed at critical locations in industrial fluid systems, such as pipe inlets and outlets, near valves, and on heat exchanger surfaces. These sensors can automatically switch sensing modes based on system operating status and environmental changes. For example, when a drastic change in fluid velocity is detected, it automatically switches to a high-frequency, high-precision velocity sensing mode.

[0036] The sensor incorporates a miniature intelligent chip to perform preliminary edge computing on the collected data. Specifically, the data is first filtered to remove noise caused by environmental interference. Then, feature extraction is performed to extract key features such as pressure change trends and temperature fluctuation frequencies, reducing the burden on subsequent data processing modules. The sampling frequency f is determined by the formula... Determined. Here, ΔL is the minimum resolvable distance of the sensor, which needs to be determined based on the specific sensor model; v max γ is the maximum flow velocity of the fluid, which can be obtained through preliminary experiments or theoretical calculations; γ is the density change influence coefficient, determined based on the fluid characteristics and experimental data; Δρ is the density change, obtained by real-time measurement from the sensor; ρ base Base density is the density of a fluid under standard conditions.

[0037] (II) Data Transmission Module

[0038] This system employs multi-path adaptive transmission technology based on Software-Defined Networking (SDN). Multiple communication paths are configured within the system, and parameters such as bandwidth, latency, and packet loss rate are monitored in real time for each path. When congestion or a failure occurs on a path, the system can quickly respond according to a formula... Calculate the optimal transmission time and dynamically adjust the data transmission path and bandwidth allocation.

[0039] Data is encrypted using blockchain-based encryption algorithms to ensure its security and immutability during transmission. Simultaneously, a Reed-Solomon code (RS code) error correction mechanism is introduced to automatically correct errors in data transmission, improving data transmission reliability.

[0040] (III) Data Processing and Analysis Module

[0041] A hybrid computing architecture based on deep learning and multiphysics models is constructed. This architecture combines deep learning models such as Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) with traditional numerical computation methods such as the finite element method and the finite difference method. CNNs are used to extract spatial features of the data; for example, when analyzing the distribution of fluids in a pipe, they can accurately identify the spatial variation patterns of parameters such as pressure and flow velocity at different locations. RNNs are used to process the time-series features of the data, such as analyzing the changing trends of fluid parameters over time.

[0042] A reinforcement learning algorithm is introduced to optimize the parameters of a multi-field coupled model in real time. This is achieved through the formula... Update the model parameters θ. Here, α is the learning rate, controlling the step size for parameter updates; A(s, a) is the advantage function, representing the advantage of taking action a relative to the average policy in state s; π θ (a|s) is the policy function, which describes the probability of taking action a in state s.

[0043] (iv) Simulation Module

[0044] The system employs real-time interactive simulation technology based on Virtual Reality (VR) / Augmented Reality (AR). Users can immerse themselves in the simulation process of industrial fluid systems using VR / AR devices such as VR headsets and AR glasses. During the simulation, users can adjust simulation parameters in real time, such as boundary conditions and physical property parameters. The system can update the simulation results in real time and display them to the user in a 3D visualization format.

[0045] Physically based rendering (PBR) technology is introduced to make simulation results more realistic and intuitive. Simultaneously, dynamic mesh adaptive technology is employed to automatically adjust the mesh density and shape according to the fluid flow state. (Mesh density) density From the formula Determined. Here, ΔV is the rate of change of fluid volume, reflecting the expansion or contraction of the fluid in space; The magnitude of the velocity gradient represents the degree of change in fluid velocity in space.

[0046] (V) Feedback Control Module

[0047] An intelligent optimization control strategy based on fuzzy adaptive model predictive control (FAMPC) is adopted. Combining the flexibility of fuzzy control and the optimization capability of model predictive control, the control parameters are adaptively adjusted according to the real-time operating status and simulation results of the industrial fluid system.

[0048] A genetic algorithm is introduced to globally optimize the parameters of FAMPC. This is achieved through the formula... Calculate the optimal control input Among them, y k+i|k The predicted output is the output value at a future time predicted based on the system model; y ref，k+i The reference output is the user-defined expected output value; Q and R are weight matrices used to adjust the output error and control the weights of input changes.

[0049] (vi) Fault Diagnosis and Early Warning Module

[0050] A fault diagnosis method based on the fusion of Deep Belief Networks (DBNs) and evidence theory is adopted. DBNs perform feature extraction and fault mode recognition on the collected data, learning data characteristics under different fault modes through training on a large amount of historical data. Evidence theory is used to fuse the diagnostic results of different fault characteristics, improving the accuracy and reliability of fault diagnosis.

[0051] A fault early warning mechanism based on time series analysis is introduced to predict the probability and timing of fault occurrence by analyzing historical and real-time data. This is achieved through formulas... Calculate the probability of failure P fault Among them, w i The weight of the i-th fault feature is determined based on its importance to fault diagnosis; D i λ represents the diagnostic result for the i-th fault feature, taking a value of 0 or 1, indicating whether the fault feature exists; i is the attenuation coefficient for the i-th fault feature, reflecting the degree of attenuation of the fault feature over time; t is time. When the probability of a fault occurrence exceeds a preset threshold, a warning signal is issued promptly, along with a detailed fault diagnosis report and solution suggestions.

[0052] (vii) Database Management Module

[0053] A hybrid storage architecture based on graph databases and distributed file systems is adopted. The graph database is used to store the topology and multi-field coupling relationships of the industrial fluid system, such as the connection relationships between pipes and the interaction relationships between different physical fields. The distributed file system is used to store large amounts of acquired data, simulation results, and control commands.

[0054] By introducing semantic search-based intelligent query technology, users can describe their query needs in natural language, and the system can automatically parse the query and return relevant results. (Using formulas...) Calculate the score of the query results. result Among them, w i The weight of the i-th query keyword is determined based on the keyword's importance; Similarity i The similarity between the query keywords and the results is calculated using a text matching algorithm.

[0055] (viii) Remote monitoring and management module

[0056] The system adopts an architecture that integrates cloud computing and edge computing. Some data processing and analysis tasks are distributed to edge nodes, reducing data transmission latency and improving system real-time performance. For example, data collected at edge devices in industrial settings undergoes preliminary processing, with only critical information transmitted to the cloud. Simultaneously, substantial storage and computing resources are deployed in the cloud to achieve centralized data management and sharing.

[0057] Users can remotely access the system via mobile devices, web browsers, and other methods to view the real-time operating status, simulation results, and control commands of the industrial fluid system. A blockchain-based identity authentication and authorization mechanism ensures the security and reliability of remote access.

[0058] III. Data Representation and Interpretation

[0059]

[0060] The data in the table above clearly demonstrates the significant advantages of this application's system compared to traditional systems. Regarding data acquisition accuracy, this system employs an adaptive multimodal sensor array and edge computing technology, enabling more accurate capture of multi-field information from industrial fluid systems, thus significantly improving acquisition accuracy. The improved data transmission success rate is attributed to SDN and blockchain-based transmission technologies, effectively ensuring secure and reliable data transmission. The substantial reduction in the error between simulation results and actual data is due to the adoption of advanced deep learning and multiphysics coupling simulation technologies, making the simulation more closely resemble real-world conditions. The improved control accuracy is a result of intelligent optimization control strategies, enabling precise adjustment of control parameters based on the system's real-time status. The significantly increased fault warning time allows enterprises more time to take preventative measures, reduce losses, and greatly improve the safety and reliability of industrial fluid systems.

[0061] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An industrial fluid mechanics multi-field coupling real-time monitoring and feedback simulation system, characterized in that, Comprise: Data acquisition module: the sensor array with adaptive sensing capability dynamically adjusts the acquisition frequency and working mode of the sensor, automatically switches the sensing mode according to different physical fields, and adjusts the acquisition frequency through the formula is determined, wherein ΔL is the minimum resolvable distance of the sensor, v max is the maximum flow rate of the fluid, γ is the density change influence coefficient, Δρ is the density change, ρ base is the base density; at the same time, the built-in chip of the sensor realizes preliminary edge computing on the collected data to remove noise and interference, and performs feature extraction; Data transmission module: using adaptive transmission technology based on software defined network (SDN) to monitor the bandwidth, delay and packet loss rate of each communication path in real time, and dynamically adjust the data transmission path and bandwidth allocation according to the parameters; At the same time, the data is encrypted based on encryption algorithm, and the data error correction code mechanism is introduced in the data transmission process, and the formula Calculate the optimal transmission time T optimal , wherein Data size is the data size, Bandwidth i is the bandwidth of the ith path, Loss i is the packet loss rate of the ith path; Data processing and analysis module: a hybrid computing architecture based on deep learning and multi-physical field model is constructed, which combines convolutional neural network (CNN), recurrent neural network (RNN) learning model, and finite element method and finite difference method to process and analyze multi-field coupling data; the reinforcement learning algorithm is introduced to optimize the parameters of the multi-field coupling model in real time, and the formula is updated to update the model parameters θ, where α is the learning rate, A(s, a) is the advantage function, π θ (a|s) is the policy function.

2. The industrial fluid mechanics multi-field coupling real-time monitoring and feedback simulation system according to claim 1, characterized in that, Also include: Simulation module: adopt real-time interactive simulation technology based on virtual reality and augmented reality, users participate in the simulation process of industrial fluid system through the device, real-time adjust the simulation parameters, the system real-time update simulation results and through 3D visualization form show to the user, at the same time, introduce the rendering PBR technology based on physics to realize that the simulation result conforms to the actual situation; In the simulation process, adopt dynamic grid adaptive technology, according to the flow state of fluid, automatically adjust the density and shape of grid, through formula Determine the grid density Mesh density , wherein AV is the volume change rate of fluid, is the modulus of velocity gradient; Feedback control module: an optimization control strategy based on fuzzy adaptive model predictive control (FAMPC) is proposed, which can adjust the control parameters according to the real-time running state and simulation results of the industrial fluid system; at the same time, genetic algorithm is introduced to optimize the parameters of FAMPC, and the optimal control input is calculated by the formula Calculate the optimal control input Where y k+i|k is the predicted output, y ref，k+i is the reference output, and Q and R are weight matrices; Fault diagnosis and early warning module: adopts the fault diagnosis method based on deep belief network DBN and evidence theory fusion, and introduces the fault early warning mechanism based on time series analysis, predicts the probability and time of fault occurrence through the analysis of historical data and real-time data; The failure occurrence probability P is calculated by the formula fault where w i is the weight of the i-th failure feature, D i is the diagnostic result of the i-th failure feature, λ i is the attenuation coefficient of the i-th failure feature, and t is time.​ When the probability of fault occurrence exceeds the preset threshold, an early warning signal is sent, and a fault diagnosis report and solution suggestion are provided.

3. The industrial fluid mechanics multi-field coupling real-time monitoring and feedback simulation system according to claim 1, characterized in that, Also include: Database management module: a hybrid storage architecture based on graph database and distributed file system is adopted, and query technology based on semantic search is introduced. Users query requirements through natural language description, and the system automatically analyzes the query statement and obtains the result. Through formula Score of calculating query result result , wherein w i is the weight of the ith query keyword, and Similarity i is the similarity between the query keyword and the result.

4. The industrial fluid mechanics multi-field coupling real-time monitoring and feedback simulation system according to claim 1, characterized in that, The sensor array of the data acquisition module adopts a bionics design concept, and the sensor array is provided with self-repairing capability. When a sensor fails, the sensing mode and data acquisition strategy are automatically adjusted, and the self-repairing capability S of the sensor is calculated by a formula S = S0 - (S0 - S1) * N repair , wherein S0 original is the original sensitivity, S1 fault is the sensitivity after failure, μ is a compensation coefficient of the redundant sensor, N redundant is the number of redundant sensors, and N total is the total number of sensors.

5. The industrial fluid mechanics multi-field coupling real-time monitoring and feedback simulation system according to claim 1, characterized in that, The data transmission module adopts a network slicing technology based on software-defined network (SDN), divides network resources into virtual network slices, each slice is customized and configured according to different service requirements and quality of service requirements, and introduces a network slice management mechanism based on distributed ledger technology, and the quality of service (QoS) of the network slice is calculated by the formula QoS = w1Q1 + w2Q2 + … + w iQi + … + w nQn slice where w i is the weight of the i th quality of service index, and QoS i is the value of the i th quality of service index.

6. The industrial fluid mechanics multi-field coupling real-time monitoring and feedback simulation system according to claim 1, characterized in that, The data processing and analysis module adopts a hybrid programming model based on a heterogeneous computing platform, combines the advantages of different computing devices such as CPU, GPU and FPGA, realizes the solution of multi-field coupling model and the calculation of learning algorithm, adopts OpenMP, CUDA and OpenCL parallel programming technology for parallel processing of computing tasks, and introduces a resource management mechanism based on task scheduling algorithm to dynamically allocate computing resources according to the characteristics of computing tasks and the performance of computing devices. The resource utilization efficiency Efficiency is calculated by the formula Efficiency = Task resource where Task completed is the number of completed computing tasks, and Resource used is the amount of computing resources used.

7. The industrial fluid mechanics multi-field coupling real-time monitoring and feedback simulation system according to claim 2, characterized in that, The simulation module adopts a collaborative simulation technology based on multi-scale modeling and multi-physical field coupling, realizes the coupling of industrial fluid system model and fluid molecular model, considers the influence of fluid structure and molecular interaction on flow characteristics, and realizes real simulation by coupling fluid mechanics, electromagnetism and thermodynamics. In the collaborative simulation process, distributed computing technology based on the Message Passing Interface (MPI) is used to improve simulation efficiency; through formulas Error in calculating simulation results simulation , where Result simulation The simulation results are shown in the Result column. experiment These are the experimental results.

8. The industrial fluid mechanics multi-field coupling real-time monitoring and feedback simulation system according to claim 2, characterized in that, The feedback control module adopts a distributed collaborative control strategy based on a multi-agent system (MAS), each control unit in the industrial fluid system is abstracted as an agent, each agent is provided with autonomous decision-making and collaborative processing capabilities; through information exchange and cooperation between the agents, distributed collaborative control of the industrial fluid system is realized; meanwhile, an agent decision-making mechanism based on game theory is introduced, the payoff of the agent is calculated by the formula Payoff = w1Utility1 + w2Utility2 + … + wNUtilityN agent where w i is the weight of the i th utility index, Utility i is the value of the i th utility index.

9. The industrial fluid mechanics multi-field coupling real-time monitoring and feedback simulation system according to claim 1, characterized in that, Also include: Remote monitoring and management module: adopts a framework based on cloud computing and edge computing fusion, and allocates part of the data processing and analysis tasks to the edge nodes for processing; At the same time, the storage and computing resources are transmitted to the cloud server to realize centralized management and sharing of data; users can remotely access the system through mobile terminals and web browsers to real-time view the running state of industrial fluid system, simulation results and control instruction information.

10. The industrial fluid mechanics multi-field coupling real-time monitoring and feedback simulation system according to claim 1, characterized in that, The system adopts modular design idea, realizes independence and scalability among modules; at the same time, introduces system integration technology based on micro-service architecture, encapsulates each module as an independent micro-service, and interacts through communication protocol; at the same time, the system provides standardized interface and open API, realizes integration and data sharing with industry system.

Images