Reverse flow valve based on early warning control of digital twinborn technology

Abstract

The invention discloses a reflux valve based on early warning control of a digital twinborn technology. The reflux valve is suitable for precise control and safety guarantee of one-way flow of fluid. The reflux valve comprises a valve body and a valve clack assembly, and a multi-source sensor is used for collecting key data such as fluid flow direction and water hammer impact characteristics in real time; the edge control unit is used for processing sensing data, monitoring countercurrent and water hammer risks, executing quick valve closing operation in an emergency, and communicating with the central processing unit at the same time; the central processing unit establishes and updates a dynamic characteristic model of the reflux valve in real time through the digital twin engine, wherein the dynamic characteristic model comprises a valve clack opening characteristic, a fluid pressure characteristic and a water hammer characteristic; the intelligent analysis early warning engine identifies the risk level and the development trend based on a deep learning algorithm, and generates an early warning signal; the self-adaptive control engine optimizes valve clack closing parameters in a control strategy library in combination with the risk signals and the dynamic data; according to the technical scheme, the intelligent level and operation reliability of the reflux valve are effectively improved, and the reflux valve can be widely applied to the field of industrial fluid control.

Description

Technical Field

[0001] The present invention relates to the technical field of reverse flow valves, and in particular to a reverse flow valve with early warning control based on digital twin technology. Background Art

[0002] In the prior art, reverse flow valves are widely used in various industrial pipeline systems to ensure the unidirectional flow of fluids and prevent the occurrence of reverse flow. Traditional reverse flow valves usually achieve automatic fluid isolation through mechanical structures, and some improved reverse flow valves introduce sensors and automatic control devices to improve performance. These reverse flow valves can monitor basic parameters such as fluid pressure and flow rate, and perform corresponding adjustment operations when abnormalities are detected, thereby mitigating water hammer impact and reverse flow risks in pipelines to a certain extent.

[0003] However, the existing reverse flow valve technology has the following problems: first, the single data collected by the sensor is often insufficient to fully reflect the complex fluid environment, resulting in inaccurate predictions of reverse flow and water hammer; second, the control logic is mostly based on fixed preset rules, which makes it difficult to dynamically adapt to changes in operating conditions, which may lead to a mismatch between the control strategy and the actual situation; finally, due to the lack of intelligent data analysis and real-time optimization methods, the existing system is less efficient when dealing with complex working conditions, which may cause excessive wear of the valve body and pipeline or operational delays.

[0004] Based on this, it is necessary to develop a reverse flow valve based on early warning control of digital twin technology. Summary of the invention

[0005] The present application provides a reverse flow valve based on early warning control of digital twin technology to improve the intelligence level and operational reliability of the reverse flow valve.

[0006] The present application provides a reverse flow valve based on early warning control of digital twin technology, comprising: A valve body, wherein a valve flap assembly is disposed in the valve body for controlling the unidirectional flow of a fluid, wherein the valve flap assembly comprises a valve flap and a valve seat; A multi-source sensor is arranged on the valve body, and the multi-source sensor includes a differential pressure sensor arranged before and after the valve disc, which is used to detect the flow direction and pressure difference of the fluid; an angle sensor arranged on the valve disc, which is used to detect the opening angle of the valve disc; a stress sensor arranged at the valve seat, which is used to detect the stress state of the valve seat sealing surface; and a vibration sensor arranged on the valve body, which is used to detect the water hammer impact characteristics; An electric actuator, disposed on the valve body, for quickly closing the valve disc when reverse flow is detected; an edge control unit, mounted on the valve body, for receiving the reverse flow valve operation data collected by the multi-source sensor; monitoring the reverse flow state and water hammer risk through the reverse flow valve operation data, and performing rapid valve closing control in an emergency; transmitting the reverse flow valve operation data collected by the multi-source sensor to the central processing unit; receiving the control command sent by the central processing unit, and adjusting the reverse flow valve through the electric actuator according to the control command; A central processing unit, the central processing unit is connected to one or more edge control units of the reverse flow valve via a communication network, comprising: A digital twin engine, configured to receive the operating data transmitted by the edge control unit; establish a digital twin model of the reverse flow valve according to the operating data; continuously update the digital twin model according to the real-time operating data to generate dynamic characteristic data including valve flap opening characteristics, fluid pressure distribution and water hammer characteristics; and transmit the dynamic characteristic data to an intelligent analysis and early warning engine; An intelligent analysis and warning engine, for receiving the dynamic characteristic data, analyzing the valve disc action characteristics and fluid state characteristics through a deep learning algorithm, identifying the risk of backflow and water hammer; generating a warning signal including the risk type and level according to the identified risk characteristics; and transmitting the warning signal to the adaptive control engine; An adaptive control engine is used to receive the dynamic characteristic data and the warning signal; based on a preset control strategy library, a basic control strategy is selected according to the risk type and level in the warning signal, wherein the control strategy library stores corresponding valve flap closing parameter templates for different risk levels; in combination with the fluid pressure distribution and water hammer characteristics in the dynamic characteristic data, the closing parameters in the basic control strategy are dynamically optimized to generate an optimal valve flap closing curve, wherein the closing curve defines the segmented speed and timing of the valve flap from the start of closure to full closure; the closing curve is converted into a control instruction of an actuator, wherein the control instruction includes a speed instruction, a position instruction and a timing instruction of the actuator; and the control instruction is sent to a corresponding edge control unit.

[0007] This application has the following beneficial technical effects: (1) Through multi-source sensors, the differential pressure, valve disc angle, sealing surface stress and vibration characteristics of the fluid are collected, and the dynamic characteristic data is updated in real time with the digital twin model, so as to achieve high-precision monitoring and prediction of the risk of reverse flow and water hammer, and significantly improve the operating reliability of the reverse flow valve. (2) Based on the warning signal and dynamic characteristic data, the adaptive control engine can dynamically adjust the valve disc closing parameters and generate the optimal closing curve, thereby providing accurate and fast response under different risk levels, reducing the wear of the valve body and pipeline, and extending the service life of the equipment. (3) The intelligent analysis and warning engine analyzes the valve disc action and fluid state characteristics through deep learning algorithms, matches them with the fault mode library, identifies the fault type and risk level in real time, and provides early warning signals to effectively avoid safety hazards caused by sudden faults. (4) Through the collaborative work of the edge control unit and the central processing unit, the distributed management of multiple reverse flow valves is realized in combination with the communication network, which enhances the flexibility and scalability of the system and adapts to complex industrial application environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 This is a schematic diagram of a reverse flow valve based on digital twin technology early warning control provided in the first embodiment of the present application. DETAILED DESCRIPTION

[0009] Many specific details are described in the following description to facilitate a full understanding of the present application. However, the present application can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the connotation of the present application, so the present application is not limited by the specific implementation disclosed below.

[0010] The first embodiment of the present application provides a reverse flow valve based on early warning control of digital twin technology. Figure 1 , which is a schematic diagram of the first embodiment of the present application. Figure 1 The first embodiment of the present application provides a reverse flow valve based on digital twin technology early warning control and is described in detail.

[0011] The ergonomic keyboard capable of automatically adjusting height and angle includes a valve body 101 , a multi-source sensor 102 , an electric actuator 103 , an edge control unit 104 and a central processing unit 105 .

[0012] A valve body 101 is provided with a valve flap assembly for controlling the unidirectional flow of a fluid. The valve flap assembly includes a valve flap and a valve seat.

[0013] The valve body 101 is the core structural component of the reverse flow valve based on the early warning control of the digital twin technology, which is used to contain the fluid and provide the working environment of the internal mechanical components (such as the valve disc assembly), and at the same time provides a platform for the installation and operation of the multi-source sensor 102, the electric actuator 103, the edge control unit 104 and its supporting systems. The outer shell of the valve body is usually made of high-strength corrosion-resistant materials, such as stainless steel or special alloys, to ensure its long-term stability and reliability in high-pressure, high-temperature and corrosive fluid environments.

[0014] The internal cavity of the valve body has a precisely designed fluid channel, the geometry of which supports unidirectional flow of the fluid and optimizes fluid dynamics. The inlet and outlet of the channel are set at both ends of the valve body, and the cross-sectional design in the channel is optimized according to the flow range of the fluid and the system pressure characteristics to reduce fluid resistance and turbulence. The installation interface of the valve body usually includes flange connection, threaded interface or welding interface to meet the connection requirements of different pipeline systems.

[0015] A disc assembly is installed inside the valve body. The assembly consists of a disc and a seat, which is responsible for realizing the one-way control function of the fluid. The valve seat is usually fixed to the downstream port of the valve body to provide a sealing surface, which cooperates with the disc to form an effective sealing state to prevent the backflow of the fluid. The sealing surface of the valve seat is usually made of wear-resistant materials, such as hardened stainless steel or coated with a special ceramic layer to extend its service life and improve its erosion resistance. The disc can rotate around a fixed axis inside the valve body, or move along a preset linear trajectory, thereby switching between opening and closing to adjust the flow state of the fluid.

[0016] The outer wall of the valve body is designed with mounting holes for multiple source sensors, including special mounting seats for differential pressure sensors, angle sensors, stress sensors and vibration sensors. The positions of the mounting holes are precisely designed to ensure that various sensors can accurately collect data on the flow status of the fluid and the operating status of the mechanical structure. For example, the differential pressure sensor is usually installed on the channel wall near the inlet and outlet to capture the upstream and downstream fluid pressure difference; the angle sensor is fastened near the valve disc bearing to monitor the rotation angle of the valve disc in real time; the stress sensor is embedded in the valve seat area to detect the stress state of the sealing surface; and the vibration sensor is arranged in the key part of the valve body shell to capture the impact characteristics of water hammer.

[0017] The top or side wall of the valve body is also provided with an interface for installing an electric actuator, which usually includes a mechanical fixing seat and a cable lead-out interface. This interface design enables the electric actuator to directly drive the valve disc assembly, ensuring that the valve disc can be quickly closed in abnormal situations to achieve effective control of the fluid.

[0018] All connections and component designs inside and outside the valve body must comply with relevant industrial standards and be precisely calibrated and sealed to ensure that the overall performance of the valve body meets the expected functional requirements and can withstand the operating environment under actual working conditions.

[0019] The multi-source sensor 102 is arranged on the valve body, and the multi-source sensor includes a differential pressure sensor arranged before and after the valve disc, which is used to detect the flow direction and pressure difference of the fluid; an angle sensor arranged on the valve disc, which is used to detect the opening angle of the valve disc; a stress sensor arranged at the valve seat, which is used to detect the stress state of the valve seat sealing surface; and a vibration sensor arranged on the valve body, which is used to detect the water hammer impact characteristics.

[0020] The multi-source sensor 102 is a key component for real-time acquisition of the reverse flow valve operating status data in this embodiment. Its function is to comprehensively monitor the dynamic characteristics of the fluid and the mechanical behavior of the valve body through the coordinated work of multiple types of sensors, and provide basic data support for subsequent analysis and control. The layout, function and operating characteristics of each sensor are precisely designed to ensure the comprehensiveness, accuracy and real-time nature of data acquisition.

[0021] The differential pressure sensor is installed at the upstream and downstream channel walls of the valve disc assembly to measure the fluid pressure values ​​at the inlet and outlet respectively, so as to calculate the pressure difference and flow direction of the fluid. The differential pressure sensor usually uses a high-sensitivity, high-pressure-resistant capacitive or piezoresistive sensor. When installed, it needs to be fixed to the valve body wall through a sealing flange or threaded interface, and a corrosion-resistant medium diaphragm is used to avoid direct contact with the fluid. The collected pressure signal is output to the edge control unit in real time through the electrical signal conversion module inside the sensor to monitor the flow direction of the fluid and abnormal pressure fluctuations.

[0022] The angle sensor is mounted near the rotating shaft or linear motion device of the valve disc to monitor the opening angle of the valve disc in real time. The sensor usually adopts a non-contact Hall effect sensor or an optical encoder, and its installation angle is precisely calibrated to ensure that the collected data can accurately reflect the dynamic position of the valve disc. The angle signal output by the sensor is used to analyze the operating status of the valve disc, such as the opening amplitude and closing speed, so as to provide support for the precise control of the valve disc movement.

[0023] The stress sensor is embedded in the sealing surface of the valve seat or in the supporting structure near it to detect the stress state of the sealing surface in real time. The sensor usually adopts a thin film resistance strain gauge or a piezoelectric stress sensor, and its installation needs to adopt an embedded structure to avoid the impact of fluid impact or mechanical interference on the data. The stress sensor can capture the force distribution of the sealing surface under different operating conditions and promptly detect the degradation of sealing performance caused by fluid abnormality or structural wear.

[0024] The vibration sensor is placed at a key position of the valve body, such as near the disc support bearing or the fluid impact point, to detect the water hammer impact characteristics. The sensor usually uses a piezoelectric or MEMS vibration sensor with a high-frequency response to identify the possible water hammer phenomenon in the fluid by collecting the vibration spectrum of the valve body. The sensor needs to be in close contact with the valve body to avoid attenuation or interference of the vibration signal.

[0025] Multi-source sensors transmit the collected signals to the edge control unit in real time through a preset communication interface. The installation position, direction and connection method of the sensors must comply with standardized design principles to ensure the stability and consistency of sensor data.

[0026] The electric actuator 103 is disposed on the valve body and is used to quickly close the valve disc when reverse flow is detected.

[0027] The electric actuator 103 is a key component in the reverse flow valve for realizing fast disc action. Its main function is to quickly drive the disc to close when reverse flow or water hammer risk is detected, thereby effectively protecting the pipeline system from damage. The design of this mechanism must meet the requirements of high response speed, high control accuracy and long-term operation reliability.

[0028] Electric actuators usually include motors, reducers, drive rods, position feedback devices and their control circuits. The motor can be a DC servo motor or a stepper motor to ensure sufficient torque and high-precision position control capabilities under complex operating conditions. The motor is connected to the drive rod through a reducer, which converts the high-speed rotary motion output by the motor into a low-speed, high-torque motion suitable for driving the valve disc, ensuring that the actuator can complete the action from fully open to fully closed or reverse in a short time.

[0029] The drive rod is designed as a high-strength metal rod, one end of which is connected to the motor output shaft, and the other end directly acts on the valve disc or transmits torque to the valve disc through a fulcrum structure. The length and stroke of the drive rod are precisely designed according to the size of the valve body and the opening and closing angle of the valve disc to ensure the displacement accuracy and reliability of each action.

[0030] The position feedback device monitors the displacement or angle of the drive rod in real time through sensors (such as encoders or Hall sensors) and provides the control unit with information about the current position of the valve disc. The feedback data is compared with the control instructions in a closed loop to ensure that the actuator's actions are always completed at the predetermined speed, position, and timing. The accuracy of the position feedback device directly affects the accuracy of the actuator's actions, so its selection must meet the requirements of high resolution and low latency.

[0031] The control circuit of the electric actuator includes a drive module and a signal processing module. The drive module is responsible for receiving speed, position and timing instructions from the edge control unit to accurately control the speed, direction and stop position of the motor; the signal processing module integrates the signal from the position feedback device and transmits it to the edge control unit to update the real-time status information of the valve disc. The control circuit usually integrates overload protection and abnormal self-detection functions to improve the operating safety of the actuator.

[0032] The actuator is usually installed outside the valve body and fixed to the valve body through a mechanical interface (such as a flange, bolts or locking structure). During installation, it is necessary to ensure that the movement direction of the drive rod is completely aligned with the movement direction of the valve disc to avoid mechanical stagnation or error accumulation during operation. In order to prevent the external environment from affecting the actuator during operation, its housing must have waterproof, dustproof and corrosion-resistant properties and meet the safety standards in relevant industrial environments (such as IP rating certification).

[0033] Through the above design and installation, the electric actuator can be ensured to achieve efficient and reliable operation under complex working conditions, and through its coordinated work with the edge control unit, it provides a solid guarantee for the rapid closing of the valve disc and the safe operation of the pipeline system. Combined with actual application requirements, the motor power, drive rod stroke and control parameters can be flexibly adjusted to adapt to the operating conditions of different pipeline systems.

[0034] The edge control unit 104 is installed on the valve body and is used to receive the reverse flow valve operation data collected by the multi-source sensor; monitor the reverse flow state and water hammer risk through the reverse flow valve operation data, and perform rapid valve closing control in an emergency; transmit the reverse flow valve operation data collected by the multi-source sensor to the central processing unit; receive the control instructions sent by the central processing unit, and adjust the reverse flow valve through the electric actuator according to the control instructions.

[0035] The edge control unit 104 is the core component responsible for real-time data processing and rapid response control in the reverse flow valve. Its main function is to receive the operating data collected by the multi-source sensor 102, analyze the data to determine the reverse flow state and water hammer risk, and directly trigger the rapid valve closing operation in an emergency. At the same time, the edge control unit also undertakes the data interaction function with the central processing unit 105, uploads the locally collected and processed data, and receives the control instructions issued by the central processing unit to implement higher-level optimization and regulation.

[0036] The edge control unit usually includes a high-performance embedded processor, a data storage module, a communication module, and an interface circuit. The embedded processor has real-time multi-tasking capabilities and can quickly complete logical operations from sensor data acquisition to valve closing. The original signal transmitted by the multi-source sensor is filtered and signal conditioned by the interface circuit before entering the processor for data fusion and analysis. The processor uses a built-in analysis algorithm to determine whether there is a risk of backflow or water hammer. For example, it calculates whether the fluid flow direction is abnormal based on the differential pressure sensor data, identifies the water hammer impact characteristics based on the vibration sensor data, and evaluates the working status of the valve disc and sealing surface based on the data of the angle sensor and stress sensor.

[0037] In an emergency, such as when a significant water hammer impact or a serious backflow risk is detected, the edge control unit can independently perform a fast valve closing operation without waiting for instructions from the central processing unit. This fast response function is achieved by the processor directly generating a valve closing instruction and sending it to the electric actuator 103. The valve closing instruction contains parameter settings for the valve disc closing speed and action timing to ensure the accuracy and timeliness of the action.

[0038] The communication module of the edge control unit supports a variety of industrial communication protocols, such as Modbus, CAN bus or industrial Ethernet, which is used to establish a reliable communication link with the central processing unit. Through the communication module, the edge control unit can transmit the collected operating data (including sensor data such as differential pressure, vibration, stress, etc.) to the central processing unit for subsequent digital twin model updates and global optimization analysis. At the same time, the edge control unit is also responsible for receiving control instructions from the central processing unit. These instructions may include optimized valve disc closing speed curves or adjustment strategies, and convert them into specific action parameters of the electric actuator.

[0039] In order to improve the reliability of the edge control unit, its hardware design takes into account factors such as high temperature, high humidity, and strong electromagnetic interference in industrial environments. The hardware circuit adopts a reinforced design, and the shell is waterproof and dustproof, meeting the protection requirements of IP65 or higher. At the same time, the edge control unit supports software fault tolerance and hardware redundancy design to ensure normal operation in the event of a single point failure.

[0040] Through the above design, the edge control unit realizes real-time data processing, local rapid response and remote collaborative control functions.

[0041] A central processing unit 105, which is connected to one or more edge control units of the reverse flow valve via a communication network, includes: A digital twin engine, configured to receive the operating data transmitted by the edge control unit; establish a digital twin model of the reverse flow valve according to the operating data; continuously update the digital twin model according to the real-time operating data to generate dynamic characteristic data including valve flap opening characteristics, fluid pressure distribution and water hammer characteristics; and transmit the dynamic characteristic data to an intelligent analysis and early warning engine; An intelligent analysis and early warning engine is used to receive the dynamic characteristic data, analyze the valve disc action characteristics and fluid state characteristics through a deep learning algorithm, identify the risk of backflow and water hammer; match the identified risk characteristics with a preset fault mode library to determine the risk level and development trend; generate an early warning signal including the risk type and level; and transmit the early warning signal to the adaptive control engine; An adaptive control engine is used to receive the dynamic characteristic data and the warning signal; based on a preset control strategy library, a basic control strategy is selected according to the risk type and level in the warning signal, wherein the control strategy library stores corresponding valve flap closing parameter templates for different risk levels; in combination with the fluid pressure distribution and water hammer characteristics in the dynamic characteristic data, the closing parameters in the basic control strategy are dynamically optimized to generate an optimal valve flap closing curve, wherein the closing curve defines the segmented speed and timing of the valve flap from the start of closure to full closure; the closing curve is converted into a control instruction of an actuator, wherein the control instruction includes a speed instruction, a position instruction and a timing instruction of the actuator; and the control instruction is sent to a corresponding edge control unit.

[0042] The central processing unit 105 is the core component responsible for global optimization analysis and control coordination in this embodiment. Its main functions are to receive data from multiple edge control units, build and update the digital twin model in real time, conduct in-depth analysis of the operating status of the reverse flow valve, and generate optimized control instructions and send them to the corresponding edge control units to achieve dynamic optimization control on a global scale.

[0043] The central processing unit is usually composed of a high-performance computing platform, including a multi-core processor, a large-capacity storage module, and a high-bandwidth communication interface. The unit is connected to multiple edge control units through a communication network, which can realize two-way high-speed data transmission. The operating data received by the central processing unit from the edge control unit includes information such as the differential pressure, vibration characteristics, valve disc angle, and stress distribution of the fluid. After these data enter the central processing unit through the communication interface, they are automatically classified and stored in the database to provide data support for subsequent analysis and modeling.

[0044] The digital twin engine is the core functional module of the central processing unit, which is used to build a digital twin model of the reverse flow valve based on real-time operating data. The model is based on the real valve body and fluid environment, and reproduces the dynamic characteristics of the valve operation through mathematical models and physical simulations, including the opening change of the valve disc, fluid pressure distribution, water hammer characteristics, etc. The digital twin model is continuously updated during operation, and new data obtained from the edge control unit is automatically integrated into the model to maintain its high-precision mapping of the actual working conditions.

[0045] The input of the digital twin model can include a variety of sensor data transmitted in real time from the edge control unit, such as the upstream and downstream pressure difference of the differential pressure sensor, the valve disc opening of the angle sensor, the sealing surface force distribution of the stress sensor, and the water hammer impact characteristics of the vibration sensor. These data are the basis for establishing and updating the model, and can reflect the key characteristics of the reverse flow valve under different operating conditions.

[0046] In order to build a digital twin model, a method based on physical modeling combined with data-driven can be used. First, a physical simulation model is established based on the design parameters of the reverse flow valve (such as valve body geometry, material properties, and fluid properties). For example, fluid mechanics software (such as ANSYS Fluent or COMSOL) can be used to simulate the pressure distribution and flow state of the fluid when it flows through the valve, calculate the force of the valve disc at different openings, and calculate the transient pressure fluctuations caused by water hammer.

[0047] Based on the physical model, calibration and optimization are performed through sensor data. Specifically, the real-time collected data can be input into the physical model to adjust the boundary conditions or material parameters in the model to make the simulation results closer to the actual working conditions. For example, when the sensor detects that the upstream pressure increases and the downstream pressure decreases rapidly, the model can dynamically adjust the fluid density and flow rate to simulate the possible water hammer phenomenon.

[0048] The output of a digital twin model often includes dynamic property data, such as: The real-time opening of the valve disc and the speed curve of its closing process; The temporal and spatial distribution of fluid pressure, including upstream and downstream pressures and local pressure on the sealing surface; Vibration characteristics, such as frequency and amplitude of water hammer impacts; Predicted failure characteristics, such as abnormal sealing surface force or valve disc movement hysteresis.

[0049] As an auxiliary module of the digital twin engine, the intelligent analysis and early warning engine receives and analyzes the dynamic characteristic data generated in the digital twin model. The engine uses a deep learning algorithm to comprehensively analyze the valve disc action characteristics, fluid flow status and other key parameters to identify possible risk characteristics, such as backflow risk, water hammer impact and abnormal sealing surface stress. The identified risk characteristics are matched with the preset fault mode library to determine their type, risk level and possible development trend. The analysis results are output in the form of early warning signals to provide input for the adaptive control engine.

[0050] In order to implement the deep learning algorithm in the intelligent analysis and early warning engine, it is necessary to extract key parameters from the dynamic characteristic data generated by the digital twin model and use this data to train a model that can identify risk characteristics. The following is a simple and feasible implementation description.

[0051] The input of the deep learning algorithm is dynamic characteristic data, which usually includes the real-time opening of the valve disc, the upstream and downstream pressure difference, the vibration frequency and the stress of the sealing surface. These data are input into the algorithm in the form of time series to capture dynamic changes and patterns. For example, multiple sets of data points within a certain time window can be used to reflect the current fluid state and mechanical behavior.

[0052] The output of the algorithm is the analysis results of risk characteristics, including risk type (such as reverse flow risk, water hammer risk or sealing abnormality), risk level (such as low risk or high risk), and possible development trend. These outputs are directly used to generate early warning signals for use by the adaptive control engine.

[0053] To implement the algorithm, a deep learning model suitable for time series data analysis can be selected, such as a long short-term memory network (LSTM) or a convolutional neural network (CNN). These models can effectively capture key patterns in time series and are suitable for analyzing the dynamic characteristics of valve operation.

[0054] The first step in building a model is data preprocessing. The dynamic feature data is organized in time series, with each time step containing a set of feature values. Then, the data needs to be normalized so that its range is consistent during model training. Next, the historical data is divided into a training set and a validation set for model learning and optimization.

[0055] The training process of the model requires data annotation, that is, marking the corresponding risk characteristics for each set of input data. For example, marking the occurrence of reverse flow risk or water hammer phenomenon in a certain period of time in the known operating data. Through these annotations, the model can learn the relationship between input data and risks.

[0056] After the model training is completed, it can be deployed in the intelligent analysis and early warning engine to receive the input data generated by the digital twin model in real time and output the risk analysis results. For example, when the valve disc opening changes rapidly and the upstream and downstream pressure differences fluctuate sharply, the model may identify high water hammer risks and output corresponding early warning signals.

[0057] In addition, the appropriate model size and amount of training data can be selected according to the complexity of the reverse flow valve system. In actual operation, the model can be continuously optimized, and the accuracy and reliability of risk identification can be further improved through the newly added operation data. In this way, the intelligent analysis and early warning engine can achieve efficient and accurate risk assessment and provide reliable support for the adaptive control engine.

[0058] Furthermore, the deep learning algorithm of the intelligent analysis and early warning engine is implemented using a pre-trained deep learning model, which includes a multi-feature decoupling module, a cross-feature correlation module, a spatiotemporal series prediction module, and a risk grading and trend prediction module; Among them, the multi-characteristic decoupling module is used to receive multi-source dynamic characteristic data from the digital twin model, and the multi-source dynamic characteristic data includes the valve flap opening change curve, the upstream and downstream pressure difference of the fluid, the sealing surface stress distribution and the vibration frequency characteristics; the output of the multi-characteristic decoupling module is an independent high-dimensional characteristic vector for each characteristic, specifically including the output of pressure fluctuation characteristics for pressure difference data, the output of vibration spectrum characteristics after Fourier transformation for vibration frequency; and the output of multi-scale time domain characteristics for valve flap opening and acceleration data; the multi-characteristic decoupling module uses a convolutional network to extract local fluctuation characteristics for the pressure difference signal, uses a Fourier domain network to extract spectrum characteristics for the vibration frequency, and uses multi-scale analysis to extract acceleration and angle change patterns for valve flap action; The input of the cross-characteristic association module is the pressure fluctuation characteristic, the vibration spectrum characteristic and the multi-scale time domain characteristic; the output of the cross-characteristic association module is the characteristic association vector and the dynamic evolution vector; the cross-characteristic association module is implemented using a multi-characteristic interaction model based on a graph neural network, performs association analysis on the decoupled characteristics, and dynamically models the coupling relationship between the characteristics; The input of the spatiotemporal sequence prediction module is the characteristic association vector and the dynamic evolution vector, and the output of the spatiotemporal sequence prediction module is the predicted characteristic trend in the time series, which includes the change prediction of water hammer impact strength, the offset prediction of sealing surface stress, and the possibility of valve flap action lag; the spatiotemporal sequence prediction module is implemented by a recursive network; The input of the risk grading and trend prediction module is the predicted characteristic trend; the risk grading and trend prediction module uses a classification regression network to perform risk grading and trend prediction on the input data, wherein the classification branch generates risk levels including backflow risk, water hammer impact and sealing abnormality according to characteristic weights and development trends; the regression branch predicts the risk intensity and possible change trends, including increased water hammer impact frequency or sealing stress offset.

[0059] The intelligent analysis and early warning engine is implemented through a pre-trained model of a deep learning algorithm. The model includes multiple modules that work together to achieve analysis and risk prediction of complex dynamic characteristics. First, the multi-characteristic decoupling module receives multi-source dynamic characteristic data from the digital twin model, such as the valve flap opening change curve, the upstream and downstream pressure difference of the fluid, the sealing surface stress distribution, and the vibration frequency characteristics. The module extracts the pressure fluctuation characteristics through a convolutional network, the vibration spectrum characteristics through a Fourier domain network, and the acceleration and angle change patterns in the valve flap action through multi-scale analysis, thereby outputting independent high-dimensional characteristic vectors for each characteristic.

[0060] The cross-feature correlation module then performs correlation analysis on these decoupled feature vectors. Through a multi-feature interaction model based on a graph neural network, the module dynamically models the coupling relationship between features, generates feature correlation vectors and dynamic evolution vectors, and reveals complex interaction patterns between different features, such as the correlation between pressure fluctuations and vibration characteristics.

[0061] Next, the spatiotemporal series prediction module receives the above characteristic association vectors and dynamic evolution vectors, and uses a recursive network to analyze the characteristic trends in the time series, including the change in water hammer impact intensity, the offset direction of the sealing surface stress, and the possible hysteresis of the valve disc action. These prediction trends provide an important basis for further risk assessment.

[0062] Finally, the risk classification and trend prediction module analyzes the predicted characteristic trends and uses the classification and regression network to classify the input data for risk and trend prediction. The classification branch generates risk levels including backflow risk, water hammer impact and sealing abnormality based on characteristic weights and trends; the regression branch predicts the risk intensity and possible development trends, such as an increase in water hammer impact frequency or a further shift in sealing stress.

[0063] Through the intelligent analysis and early warning engine, the present invention can comprehensively and efficiently identify and predict the operating risks of the reverse flow valve, and provide accurate risk signals and adjustment suggestions for subsequent control modules.

[0064] Furthermore, the spatiotemporal sequence prediction module is specifically used for: receiving the feature association vector and the dynamic evolution vector from the cross feature association module, and grouping the input data into multi-dimensional time series, wherein the grouping method is dynamically adjusted according to the time correlation in the feature association vector; The grouped time series data are processed using a recursive network. In the first stage of the recursive network, short-term dynamic characteristic patterns are extracted through unidirectional recursive units, including pressure change trends, vibration spectrum changes, and stress displacement rates. In the second stage, short-term and long-term dynamic characteristics are combined through a bidirectional recursive unit to capture the potential causal relationship between pressure and vibration characteristics; Use the temporal attention mechanism to assign weights to different time steps and dynamically enhance attention to key time points, such as abnormal enhancement of vibration spectrum or rapid fluctuation of pressure difference; The output is the predicted characteristic trend in the time series, including the changing trend of water hammer impact intensity, the offset direction of sealing surface stress and the time range of possible delay of valve disc action.

[0065] The spatiotemporal series prediction module of the present invention is designed to perform efficient time series analysis and trend prediction on complex dynamic characteristic data. The module receives characteristic association vectors and dynamic evolution vectors from the cross characteristic association module, and groups these input data into multidimensional time series according to time correlation. The grouping method can be dynamically adjusted to adapt to the time dependency changes between different characteristics in the characteristic association vector, thereby ensuring the accuracy of the analysis.

[0066] The module processes the grouped time series data through a recursive network. In the first stage, a unidirectional recursive unit is used to extract short-term dynamic characteristic patterns, which can identify pressure change trends, local changes in the vibration spectrum, and the rate of stress displacement. This stage focuses on capturing rapid changes and local characteristics.

[0067] In the second stage, the bidirectional recursive unit combines short-term and long-term dynamic characteristics to further capture the potential causal relationship between characteristics by analyzing the information of the previous and next correlations in the time series. For example, the synchronization of pressure changes and vibration spectra can be analyzed to assess the cause of water hammer risk.

[0068] In addition, the module integrates a temporal attention mechanism to enhance the focus on key time points by assigning dynamic weights to different time steps. This mechanism can identify important characteristic changes such as abnormal enhancement of vibration spectrum or rapid fluctuation of pressure difference, and highlight the importance of these key information for subsequent analysis.

[0069] Finally, the spatiotemporal series prediction module outputs the predicted characteristic trends in the time series, including the changing trend of water hammer impact intensity, the offset direction of sealing surface stress, and the possible time range of valve disc action lag. These outputs provide accurate data support for risk assessment and control decisions.

[0070] Furthermore, the risk grading and trend prediction module includes: Receive the predicted characteristic trend data from the spatiotemporal series prediction module, and first assign importance weights to each characteristic through the characteristic weight adjustment module, specifically including increasing the weights of vibration spectrum changes and pressure fluctuations to highlight the analysis of water hammer risks; Use the classification branch of the classification regression network to classify the adjusted input data, combine the predicted trend data with the characteristic weights, and generate risk types and their levels including backflow risk, water hammer impact and sealing abnormality; Use the regression branch of the classification regression network to predict the trend of the input data, and output the numerical prediction of the risk intensity and the development trend for each risk type, such as the gradual increase in the frequency of water hammer impact or the expansion of the sealing stress excursion range; The output includes comprehensive prediction results of risk type, risk level and development trend, and generates early warning signals for adaptive control engine.

[0071] The risk classification and trend prediction module is responsible for in-depth analysis of the predicted characteristic trend data and provides detailed risk assessment and trend prediction results. The module first receives the predicted characteristic trend data from the spatiotemporal series prediction module and assigns importance weights to each characteristic through the characteristic weight adjustment module. For example, for the analysis of water hammer risk, the module will appropriately increase the weights of vibration spectrum changes and pressure fluctuations according to the actual working conditions to more accurately capture the relevant risk characteristics.

[0072] The adjusted data is input into the classification branch of the classification regression network for risk grading. The classification branch combines the predicted trend data and characteristic weights to generate risk types including backflow risk, water hammer impact and sealing abnormality, and assigns a risk level (such as low, medium, and high) to each risk type. This process ensures that different types of risks can be clearly identified and provides a clear priority reference for subsequent control decisions.

[0073] The module then uses the regression branch of the classification regression network to predict the trend of the input data. The regression branch generates specific numerical predictions for each risk type, such as the frequency change of water hammer impact, the trend of the change in water hammer impact amplitude, and the trend of the expansion of the seal stress excursion range. These numerical predictions further refine the risk assessment and provide high-value input data for precise control.

[0074] The final output of the module includes comprehensive prediction results of risk type, risk level and development trend. These results are further converted into early warning signals for the adaptive control engine to guide subsequent control strategy adjustments.

[0075] Furthermore, the multi-characteristic decoupling module extracts the local fluctuation characteristics of the pressure difference signal through the following formula 1: ;

[0076] in, Represents the total number of convolution kernels used to extract local fluctuation characteristics; In time The pressure fluctuation characteristic vector, indicating the pressure fluctuation at time The pressure fluctuation characteristic vector at is the final extracted characteristic output, which is used to describe the local fluctuation pattern of the pressure difference signal. Each element of this vector corresponds to an extracted characteristic value, which is used as the output of the multi-characteristic decoupling module for analysis in subsequent modules.

[0077] It's in time The first one in the pressure difference signal sequence input at The specific value comes from the upstream and downstream pressure difference signal collected by the sensor. , As input signal, in time window Divide and process.

[0078] is the time window length, which is used to capture the short-term fluctuation characteristics of the signal; the recommended value is determined according to the signal sampling frequency. For example, when the sampling frequency is 100 Hz, it is recommended to take = 0.1 sec.

[0079] is the Gaussian weighted kernel function; is the convolution kernel weight; is the periodic weight function; is the variance of the pressure difference signal within the time window, which is used to enhance the sensitivity to local abnormal changes; Variance gain weight, used to control the contribution of the variance term. The recommended value is 0.5 to 2.0, which should be adjusted according to the noise level of the signal and the requirement for abnormal sensitivity.

[0080] It is a bias term used to balance the output. It is recommended to initialize it to zero or set it according to the data distribution.

[0081] The Gaussian weighted kernel function is implemented using the following formula 2: ;

[0082] in, It's time The signal value at taken out from the sequence.

[0083] is the convolution kernel weight, which indicates the sensitivity to a specific pattern of the signal. The recommended initial value is Gaussian random distribution, which is then optimized through training.

[0084] is the Gaussian kernel width, which determines the sensitivity of the kernel function to signal deviation. The recommended value range is 0.1 to 1.0, adjusted according to the signal noise level.

[0085] The periodic weight function is implemented using the following formula 3: ;

[0086] in, is the frequency parameter that controls the oscillation frequency of the period weight. The recommended value is a multiple of the main frequency component of the signal, for example Indicates a frequency of 50 Hz.

[0087] is the phase parameter, which is used to adjust the phase offset of the weight function. It is recommended to initialize it to a random distribution.

[0088] Furthermore, the cross-feature association module models the dynamic coupling relationship between features through the following formula 4: ;

[0089] in, Representation characteristics and Features In time The correlation weight of ; used to describe the coupling strength between two features at a specific time point. This value is directly calculated by the formula and is the output of the cross-feature correlation module, which is used for feature fusion and dynamic evolution analysis of subsequent modules.

[0090] The number of frequency layers that represent the coupling between features is the number of levels of frequency components in the formula. Each frequency layer corresponds to a specific coupling analysis range. For example, a low-frequency layer can capture slow changes between features, and a high-frequency layer can capture fast fluctuations. The recommended value is 3 to 5, depending on the complexity of the system.

[0091] For time Lower characteristics and No. Layer coupling strength represents the weight of different frequency layers. Its initial value is usually determined by training data. For example, it can be initialized to a normal distribution random value based on the frequency distribution of historical characteristic coupling relationship, and then optimized through training.

[0092] Representation characteristics and No. Layer-specific frequency components are used to capture the periodic relationship between characteristics. For example, if the system signal is mainly distributed in the low frequency range, ,in is the fundamental frequency (recommended value 10 Hz).

[0093] For Features and In time The time offset is defined as: ;

[0094] in, Yes Features signal delay or peak occurrence time. Yes Features The signal delay or peak occurrence time of these values ​​can be obtained by real-time signal processing methods such as peak detection.

[0095] is a small positive number used to avoid the denominator being zero; For Features In time The characteristic value of For Features In time The feature values ​​of a feature; these values ​​are directly output from the previous module (such as the multi-feature decoupling module) and represent the real-time data of the feature at a specific point in time.

[0096] It is the feature difference weight, which is used to amplify the nonlinear relationship between features. Its value indicates the sensitivity to the difference between features. The recommended range is 0.1 to 1.0, which is adjusted according to the complexity of system features. The initial value can be set to 0.5 and updated during training.

[0097] The following is the reference implementation code of the deep learning model: import torch import torch.nn as nn import torch.optim as optim import numpy as np # Define multi-feature decoupling modules class MultiFeatureDecomposition(nn.Module): def __init__(self, input_dim, num_kernels, kernel_width): super(MultiFeatureDecomposition, self).__init__() self.num_kernels = num_kernels self.kernel_width = kernel_width self.weights = nn.Parameter(torch.randn(num_kernels, input_dim)) self.bias = nn.Parameter(torch.zeros(num_kernels)) self.sigma = nn.Parameter(torch.ones(num_kernels) kernel_width) self.frequency_params = nn.Parameter(torch.ones(num_kernels) 2 np.pi) self.phase_params = nn.Parameter(torch.zeros(num_kernels)) def forward(self, x): # Gaussian weighted kernel function gauss_kernel = torch.exp(-torch.square(x.unsqueeze(1) - self.weights) / (2 torch.square(self.sigma))) # Periodic weight function time_indices = torch.arange(x.size(1), device=x.device).float().unsqueeze(0) periodic_weights = torch.cos(self.frequency_params.unsqueeze(1) time_indices + self.phase_params.unsqueeze(1)) # Weight the pressure fluctuation characteristics weighted_output = torch.sum(gauss_kernel periodic_weights, dim=2) + self.bias.unsqueeze(0) # Variance enhancement variance = torch.var(x, dim=1, unbiased=False).unsqueeze(1) enhanced_output = weighted_output + 0.5 variance return enhanced_output # Define the cross - feature association module class CrossFeatureAssociation(nn.Module): def __init__(self, num_features, num_frequency_layers): super(CrossFeatureAssociation, self).__init__() self.num_frequency_layers = num_frequency_layers self.alpha = nn.Parameter(torch.randn(num_features, num_features, num_frequency_layers)) self.frequency = nn.Parameter(torch.linspace(0.1, 1.0, num_frequency_layers)) self.lambda_weights = nn.Parameter(torch.ones(num_features,num_features, num_frequency_layers)) def forward(self, y_i, y_j, time_deltas): time_deltas = time_deltas.unsqueeze(-1).expand(-1, -1,self.num_frequency_layers) sine_term = torch.sin(np.pi self.frequency time_deltas) / (np.pi self.frequency time_deltas + 1e-6) nonlinear_difference = torch.log(1 + torch.abs(y_i.unsqueeze(2) - y_j.unsqueeze(2))) association = torch.sum(self.alpha sine_term (1 +self.lambda_weights nonlinear_difference), dim=2) return association # Define the spatio-temporal sequence prediction module class SpatioTemporalPrediction(nn.Module): def __init__(self, input_dim, hidden_dim): super(SpatioTemporalPrediction, self).__init__() self.rnn1 = nn.LSTM(input_dim, hidden_dim, batch_first=True) self.rnn2 = nn.LSTM(hidden_dim, hidden_dim, batch_first=True,bidirectional=True) self.attention_weights = nn.Linear(hidden_dim * 2, 1) def forward(self, x): rnn1_output, _ = self.rnn1(x) rnn2_output, _ = self.rnn2(rnn1_output) attention_scores = torch.softmax(self.attention_weights(rnn2_output), dim=1) weighted_output = torch.sum(rnn2_output * attention_scores, dim=1) return weighted_output # Define the risk classification and trend prediction module class RiskClassificationAndTrendPrediction(nn.Module): def __init__(self, input_dim, num_classes): super(RiskClassificationAndTrendPrediction, self).__init__() self.feature_weight_adjustment = nn.Linear(input_dim, input_dim) self.classification_branch = nn.Linear(input_dim, num_classes) self.regression_branch = nn.Linear(input_dim, 1) def forward(self, x): adjusted_features = torch.relu(self.feature_weight_adjustment(x)) classifications = torch.softmax(self.classification_branch(adjusted_features), dim=1) trends = self.regression_branch(adjusted_features) return classifications, trends # Integrated model implementation class IntelligentAnalysisEngine(nn.Module): def __init__(self, input_dim, num_kernels, kernel_width, num_frequency_layers, hidden_dim, num_classes): super(IntelligentAnalysisEngine, self).__init__() self.decomposition = MultiFeatureDecomposition(input_dim, num_kernels, kernel_width) self.association = CrossFeatureAssociation(num_kernels, num_frequency_layers) self.prediction = SpatioTemporalPrediction(num_kernels,hidden_dim) self.risk_module = RiskClassificationAndTrendPrediction(hidden_dim, num_classes) def forward(self, x, time_deltas): decomposed_features = self.decomposition(x) associations = self.association(decomposed_features,decomposed_features, time_deltas) temporal_predictions = self.prediction(associations) classifications, trends = self.risk_module(temporal_predictions) return classifications, trends To fully train and deploy the above deep learning model, you can follow the following process: First, collect and process multi-source dynamic characteristic data, including valve disc opening change curve, upstream and downstream pressure difference, sealing surface stress distribution and vibration frequency characteristics. Convert the raw data into a format suitable for input model, such as time series tensor, and ensure that each characteristic data has the same time alignment and sampling rate. Standardize the data and scale the characteristic values ​​to a range with a mean of 0 and a standard deviation of 1 to improve the convergence speed and stability of the model. Next, divide the data into training set, validation set and test set. It is recommended to divide them in a ratio of 80%, 10%, and 10% to ensure that each part of the data evenly covers different operating states.

[0098] During the training phase, the loss function is defined, including the cross entropy loss for the classification task and the mean square error loss for the trend prediction task. At the same time, a weight balancing strategy is designed to ensure that the contribution of the classification and regression tasks to the total loss is appropriate. Using the Adam optimizer, the initial learning rate is set to 0.001, and a learning rate scheduler is added to dynamically adjust the learning rate to adapt to the convergence of the model. The model training adopts a batch method, and the recommended batch size is 32 to ensure a balance between training efficiency and video memory usage. After each training cycle (epoch), the validation set is used to evaluate the model performance, including classification accuracy, mean square error of trend prediction, and overall loss change. When the performance of the validation set is no longer significantly improved, the training can be stopped.

[0099] After training is complete, the generalization performance of the model is evaluated using the test set to ensure that its performance on unseen data meets expectations. To deploy the model, first save the trained model parameters as a file. Next, load the model in the deployment environment, implement data preprocessing and reasoning processes, input real-time collected data into the model, and obtain risk classification and trend prediction results.

[0100] Finally, through the integrated deployment system, such as edge devices or servers, the model output is passed to the adaptive control module to achieve closed-loop control, and the model performance is continuously monitored and optimized in actual operation.

[0101] The adaptive control engine performs global optimization calculations based on early warning signals and dynamic characteristic data to select the optimal valve flap closing strategy. Specifically, the adaptive control engine selects basic strategies from the preset control strategy library according to the fault type and risk level, and optimizes the closing parameters in combination with real-time data. For example, for severe water hammer impact, the control engine may generate a nonlinear fast closing curve; in the case of mild backflow, it may choose to close slowly to reduce system impact. The optimized closing curve is converted into specific speed, position and timing instructions, which are ultimately sent to the corresponding edge control unit through the communication network, which drives the electric actuator to perform the corresponding operation.

[0102] Furthermore, the adaptive control engine is specifically used for: Receive warning signals from the intelligent analysis and warning engine, analyze risk types and risk levels, and identify priority control targets based on risk types, including adjusting the valve disc closing speed for water hammer risks and adjusting the valve seat stress state for sealing anomalies; Query the preset control strategy library and select the corresponding basic control strategy according to the risk level, including slow shutdown strategy for low risk and fast shutdown strategy for high risk; The basic control strategy is used as the initial solution and input into the next step of dynamic optimization process.

[0103] The adaptive control engine is used to receive the warning signal generated by the intelligent analysis and warning engine, and analyze it according to the risk type and risk level contained in the signal. First, the adaptive control engine classifies the risk type, such as water hammer risk or sealing abnormality, and determines the priority control target based on the classification. For example, for water hammer risk, the control target is to adjust the closing speed of the valve disc to reduce the impact of water hammer impact; while for sealing abnormality, the control target is to optimize the force state of the valve seat to ensure sealing performance.

[0104] The adaptive control engine then retrieves the basic control strategy suitable for the current risk level from the preset control strategy library. For example, in low-risk situations, a slow closing strategy can be selected to reduce wear on the valve body and sealing components; while in high-risk situations, a fast closing strategy can be selected to minimize the impact of water hammer or backflow in an emergency. Each strategy in the control strategy library predefines the specific closing speed and action sequence to ensure that reasonable basic control parameters can be provided under different risk levels.

[0105] Finally, the adaptive control engine uses the selected basic control strategy as the initial solution and inputs it into the dynamic optimization process, and further optimizes it in combination with real-time data. Through the above steps, the adaptive control engine can efficiently respond to early warning signals and achieve precise control of the valve.

[0106] Furthermore, the adaptive control engine optimizes the parameters in the basic control strategy according to the dynamic characteristic data, specifically including: Analyze the fluid pressure distribution, water hammer characteristics and valve disc opening change trend in the dynamic characteristic data, and dynamically adjust the valve disc closing speed according to the pressure change rate; Real-time adjustment of the closing curve based on the frequency and amplitude of water hammer impact, ensuring fast closing at the low amplitude stage of water hammer impact and avoiding pressure shock at the high amplitude stage; Optimize the closing sequence according to the stress state of the sealing surface to avoid excessive or uneven stress on the sealing surface due to closing too quickly.

[0107] The adaptive control engine uses dynamic characteristic data to optimize the parameters of the basic control strategy to achieve more accurate valve control. First, the fluid pressure distribution, water hammer characteristics and valve flap opening change trend in the dynamic characteristic data are used to analyze the current operating status of the system. By monitoring the pressure change rate, the adaptive control engine can adjust the closing speed of the valve flap in real time to synchronize the closing action with the pressure change to reduce the additional impact caused by pressure fluctuations.

[0108] When water hammer is detected, the adaptive control engine further analyzes the frequency and amplitude of the water hammer impact and adjusts the valve flap closing curve based on this information. In the stage of low amplitude of water hammer impact, the adaptive control engine prioritizes the fast closing strategy to reduce the continuous impact that may be caused by water hammer; in the stage of high amplitude, it controls the closing speed and timing to avoid further aggravating the pressure shock of the system.

[0109] In addition, the adaptive control engine will optimize the closing timing according to the stress state of the sealing surface to balance the force distribution of the closing action on the sealing surface. By preventing excessive or uneven local force on the sealing surface due to too fast closing, the adaptive control engine can extend the service life of the sealing surface and improve the reliability of the seal.

[0110] In this way, the adaptive control engine can dynamically adjust the closing parameters to make the valve operation more in line with the needs of real-time working conditions.

[0111] Furthermore, the adaptive control engine generates an optimal valve flap closing curve through a multi-objective optimization method, specifically including: Define multi-objective optimization goals, including minimizing the valve disc closing time, minimizing the sealing surface force during closing, and reducing the water hammer impact energy; Use a dynamic weight adjustment mechanism to assign weights to different optimization goals, including prioritizing reducing shutdown time when high risk levels are detected; The optimal closing curve is generated by an iterative optimization method, which contains the segmented speed and timing of the valve disc from starting to closing to full closing.

[0112] The adaptive control engine generates the optimal flap closing curve through a multi-objective optimization method to achieve precise control of the reverse flow valve under different risk conditions. The optimization objectives include minimizing the flap closing time, reducing the force on the sealing surface during the closing process, and reducing the impact energy of water hammer. To ensure the dynamic adaptability of the optimization objectives, the adaptive control engine uses a weight dynamic adjustment mechanism to adjust the priority of the optimization objectives according to the real-time working conditions. For example, when a high risk level is detected, more weight is assigned to the goal of reducing the closing time to curb potential risks as soon as possible; while in low-risk situations, more weight will be assigned to the goal of reducing the force on the sealing surface to extend the service life of the component.

[0113] During the optimization process, the adaptive control engine uses an iterative optimization method. By continuously adjusting parameters and verifying outputs, the adaptive control engine generates a closing curve that includes segmented speeds and timings from the start of the valve flap to full closure, and each segment is optimized to ensure the best balance of goals in different time periods. This iterative method can not only adapt to real-time dynamic characteristic data, but also generate high-precision control parameters for complex working conditions.

[0114] Through the above optimization method, the adaptive control engine can generate the optimal closing curve that combines efficiency and safety.

[0115] Furthermore, the adaptive control engine converts the optimized closing curve into control instructions for the electric actuator, specifically including: Analyze the segmented speed and timing parameters in the closing curve into speed instructions and position instructions of the electric actuator; Add redundant check parameters to each control command to ensure that command errors caused by noise or data loss during transmission can be detected and corrected; The parsed control command is sent to the edge control unit, which drives the electric actuator to perform the valve flap closing operation.

[0116] After generating the optimized closing curve, the adaptive control engine converts it into control instructions that can be directly executed by the electric actuator to achieve accurate closing of the valve disc. The segmented speed and timing parameters in the closing curve are key inputs. The adaptive control engine generates specific speed instructions and position instructions by parsing these parameters. For example, the segmented speed within a certain time period is used to set the operating speed of the electric actuator, while the timing parameters determine the position change of the valve disc at different stages.

[0117] To improve the reliability of control instructions, the adaptive control engine adds redundant check parameters to each control instruction generated. These parameters are used to detect possible errors during instruction transmission, such as anomalies caused by communication noise or data loss. Through the built-in error checking and correction mechanism, the adaptive control engine can ensure that the instructions transmitted to the edge control unit are accurate.

[0118] Finally, the adaptive control engine sends the parsed control instructions to the edge control unit. The edge control unit accurately drives the electric actuator to close the valve disc according to the received speed and position instructions. The entire process, from parsing the closing curve to completing the execution of the action, is aimed at optimizing efficiency and ensuring safety.

[0119] The hardware and software design of the central processing unit must take into account the high reliability and high efficiency requirements of the industrial environment. The hardware part adopts anti-interference design and supports redundant backup to prevent single point failures; the software part uses fault-tolerant algorithms and real-time task scheduling mechanisms to ensure stable operation of the system under complex working conditions.

[0120] Although the present application is disclosed as above in the form of a preferred embodiment, it is not intended to limit the present application. Any technical personnel in this field may make possible changes and modifications without departing from the spirit and scope of the present application. Therefore, the scope of protection of the present application shall be based on the scope defined by the claims of the present application.

Claims

1. A reverse flow valve based on early warning control of digital twin technology, characterized in that: include: A valve body, wherein a valve flap assembly is disposed in the valve body for controlling the unidirectional flow of a fluid, wherein the valve flap assembly comprises a valve flap and a valve seat; A multi-source sensor is arranged on the valve body, and the multi-source sensor includes a differential pressure sensor arranged before and after the valve disc, which is used to detect the flow direction and pressure difference of the fluid; an angle sensor arranged on the valve disc, which is used to detect the opening angle of the valve disc; a stress sensor arranged at the valve seat, which is used to detect the stress state of the valve seat sealing surface; and a vibration sensor arranged on the valve body, which is used to detect the water hammer impact characteristics; An electric actuator, disposed on the valve body, for quickly closing the valve disc when reverse flow is detected; an edge control unit, mounted on the valve body, for receiving the reverse flow valve operation data collected by the multi-source sensor; monitoring the reverse flow state and water hammer risk through the reverse flow valve operation data, and performing rapid valve closing control in an emergency; transmitting the reverse flow valve operation data collected by the multi-source sensor to the central processing unit; receiving the control command sent by the central processing unit, and adjusting the reverse flow valve through the electric actuator according to the control command; A central processing unit, the central processing unit is connected to one or more edge control units of the reverse flow valve via a communication network, comprising: A digital twin engine, configured to receive the operating data transmitted by the edge control unit; establish a digital twin model of the reverse flow valve according to the operating data; continuously update the digital twin model according to the real-time operating data to generate dynamic characteristic data including valve flap opening characteristics, fluid pressure distribution and water hammer characteristics; and transmit the dynamic characteristic data to an intelligent analysis and early warning engine; An intelligent analysis and warning engine, for receiving the dynamic characteristic data, analyzing the valve disc action characteristics and fluid state characteristics through a deep learning algorithm, identifying the risk of backflow and water hammer; generating a warning signal including the risk type and level according to the identified risk characteristics; and transmitting the warning signal to the adaptive control engine; An adaptive control engine is used to receive the dynamic characteristic data and the warning signal; based on a preset control strategy library, a basic control strategy is selected according to the risk type and level in the warning signal, wherein the control strategy library stores corresponding valve flap closing parameter templates for different risk levels; in combination with the fluid pressure distribution and water hammer characteristics in the dynamic characteristic data, the closing parameters in the basic control strategy are dynamically optimized to generate an optimal valve flap closing curve, wherein the closing curve defines the segmented speed and timing of the valve flap from the start of closure to full closure; the closing curve is converted into a control instruction of an actuator, wherein the control instruction includes a speed instruction, a position instruction and a timing instruction of the actuator; and the control instruction is sent to a corresponding edge control unit.

2. The reverse flow valve based on early warning control of digital twin technology according to claim 1 is characterized in that: The adaptive control engine is specifically used for: Receive warning signals from the intelligent analysis and warning engine, analyze risk types and risk levels, and identify priority control targets based on risk types, including adjusting the valve disc closing speed for water hammer risks and adjusting the valve seat stress state for sealing anomalies; Query the preset control strategy library and select the corresponding basic control strategy according to the risk level, including slow shutdown strategy for low risk and fast shutdown strategy for high risk; The basic control strategy is used as the initial solution and input into the next step of dynamic optimization process.

3. The reverse flow valve based on early warning control of digital twin technology according to claim 1 is characterized in that: The adaptive control engine optimizes the parameters in the basic control strategy according to the dynamic characteristic data, specifically including: Analyze the fluid pressure distribution, water hammer characteristics and valve disc opening change trend in the dynamic characteristic data, and dynamically adjust the valve disc closing speed according to the pressure change rate; Real-time adjustment of the closing curve based on the frequency and amplitude of water hammer impact, ensuring fast closing at the low amplitude stage of water hammer impact and avoiding pressure shock at the high amplitude stage; Optimize the closing sequence according to the stress state of the sealing surface to avoid excessive or uneven stress on the sealing surface due to closing too quickly.

4. The reverse flow valve based on early warning control of digital twin technology according to claim 1 is characterized in that: The adaptive control engine generates an optimal valve flap closing curve through a multi-objective optimization method, specifically including: Define multi-objective optimization goals, including minimizing the valve disc closing time, minimizing the sealing surface force during closing, and reducing the water hammer impact energy; Use a dynamic weight adjustment mechanism to assign weights to different optimization goals, including prioritizing reducing shutdown time when high risk levels are detected; The optimal closing curve is generated by an iterative optimization method, which contains the segmented speed and timing of the valve disc from starting to closing to full closing.

5. The reverse flow valve based on early warning control of digital twin technology according to claim 1 is characterized in that: The adaptive control engine converts the optimized closing curve into control instructions for the electric actuator, specifically including: Analyze the segmented speed and timing parameters in the closing curve into speed instructions and position instructions of the electric actuator; Add redundant check parameters to each control command to ensure that command errors caused by noise or data loss during transmission can be detected and corrected; The parsed control command is sent to the edge control unit, which drives the electric actuator to perform the valve flap closing operation.

6. The reverse flow valve based on early warning control of digital twin technology according to claim 1 is characterized in that: The deep learning algorithm of the intelligent analysis and early warning engine is implemented using a pre-trained deep learning model, which includes a multi-feature decoupling module, a cross-feature association module, a spatiotemporal sequence prediction module, and a risk grading and trend prediction module; Among them, the multi-characteristic decoupling module is used to receive multi-source dynamic characteristic data from the digital twin model, and the multi-source dynamic characteristic data includes the valve flap opening change curve, the upstream and downstream pressure difference of the fluid, the sealing surface stress distribution and the vibration frequency characteristics; the output of the multi-characteristic decoupling module is an independent high-dimensional characteristic vector for each characteristic, specifically including the output of pressure fluctuation characteristics for pressure difference data, the output of vibration spectrum characteristics after Fourier transformation for vibration frequency; and the output of multi-scale time domain characteristics for valve flap opening and acceleration data; the multi-characteristic decoupling module uses a convolutional network to extract local fluctuation characteristics for the pressure difference signal, uses a Fourier domain network to extract spectrum characteristics for the vibration frequency, and uses multi-scale analysis to extract acceleration and angle change patterns for valve flap action; The input of the cross-characteristic association module is the pressure fluctuation characteristic, the vibration spectrum characteristic and the multi-scale time domain characteristic; the output of the cross-characteristic association module is the characteristic association vector and the dynamic evolution vector; the cross-characteristic association module is implemented using a multi-characteristic interaction model based on a graph neural network, performs association analysis on the decoupled characteristics, and dynamically models the coupling relationship between the characteristics; The input of the spatiotemporal sequence prediction module is the characteristic association vector and the dynamic evolution vector, and the output of the spatiotemporal sequence prediction module is the predicted characteristic trend in the time series, which includes the change prediction of water hammer impact strength, the offset prediction of sealing surface stress, and the possibility of valve flap action lag; the spatiotemporal sequence prediction module is implemented by a recursive network; The input of the risk grading and trend prediction module is the predicted characteristic trend; the risk grading and trend prediction module uses a classification regression network to perform risk grading and trend prediction on the input data, wherein the classification branch generates risk levels including backflow risk, water hammer impact and sealing abnormality according to characteristic weights and development trends; the regression branch predicts the risk intensity and possible change trends, including increased water hammer impact frequency or sealing stress offset.

7. The reverse flow valve based on early warning control of digital twin technology according to claim 6 is characterized in that: The spatiotemporal sequence prediction module is specifically used for: receiving the feature association vector and the dynamic evolution vector from the cross feature association module, and grouping the input data into multi-dimensional time series, wherein the grouping method is dynamically adjusted according to the time correlation in the feature association vector; The grouped time series data are processed using a recursive network. In the first stage of the recursive network, short-term dynamic characteristic patterns are extracted through unidirectional recursive units, including pressure change trends, vibration spectrum changes, and stress displacement rates. In the second stage, short-term and long-term dynamic characteristics are combined through a bidirectional recursive unit to capture the potential causal relationship between pressure and vibration characteristics; Use the temporal attention mechanism to assign weights to different time steps and dynamically enhance attention to key time points, such as abnormal enhancement of vibration spectrum or rapid fluctuation of pressure difference; The output is the predicted characteristic trend in the time series, including the changing trend of water hammer impact intensity, the offset direction of sealing surface stress and the time range of possible delay of valve disc action.

8. The reverse flow valve based on early warning control of digital twin technology according to claim 6 is characterized in that: The risk grading and trend prediction module is specifically used for: Receive the predicted characteristic trend data from the spatiotemporal series prediction module, and first assign importance weights to each characteristic through the characteristic weight adjustment module, specifically including increasing the weights of vibration spectrum changes and pressure fluctuations to highlight the analysis of water hammer risks; Use the classification branch of the classification regression network to classify the adjusted input data, combine the predicted trend data with the characteristic weights, and generate risk types and their levels including backflow risk, water hammer impact and sealing abnormality; Use the regression branch of the classification regression network to predict the trend of the input data, and output the numerical prediction of the risk intensity and the development trend for each risk type, such as the gradual increase in the frequency of water hammer impact or the expansion of the sealing stress excursion range; The output includes comprehensive prediction results of risk type, risk level and development trend, and generates early warning signals for adaptive control engine.

9. The reverse flow valve based on early warning control of digital twin technology according to claim 6 is characterized in that: The multi-characteristic decoupling module extracts the local fluctuation characteristics of the pressure difference signal through the following formula 1: ； in, Represents the total number of convolution kernels used to extract local fluctuation characteristics; In time The pressure fluctuation characteristic vector of It's in time The first one in the pressure difference signal sequence input at values; is the time window length, which is used to capture the short-term fluctuation characteristics of the signal; is the Gaussian weighted kernel function; is the convolution kernel weight; is the periodic weight function; is the variance of the pressure difference signal within the time window, which is used to enhance the sensitivity to local abnormal changes; is the variance gain weight; is a bias term used to balance the output; the Gaussian weighted kernel function is implemented using the following formula 2: ； in, is the convolution kernel weight, is the Gaussian kernel width; the periodic weight function is implemented using the following formula 3: ； in, is the frequency parameter; is the phase parameter.

10. The reverse flow valve based on early warning control of digital twin technology according to claim 6 is characterized in that: The cross-feature association module models the dynamic coupling relationship between features through the following formula 4: ； in, Representation characteristics and Features In time The association weight of The number of frequency layers representing coupling between features; For time Lower characteristics and No. Layer coupling strength; Representation characteristics and No. layer-specific frequency content; For Features and In time The time offset of is a small positive number used to avoid the denominator being zero; For Features In time The characteristic value of For Features In time The characteristic value of is the feature difference weight, which is used to amplify the nonlinear relationship between features.

Images