An energy-saving intelligent switching valve body operation method and system

Abstract

The present invention relates to the technical field of intelligent valve body control, and discloses an energy-saving intelligent switching valve body operation method and system. The method obtains the historical operation data of the intelligent switching valve body, generates a structured feature set according to preset dimensions such as the fluid pressure fluctuation curve; constructs a multi-dimensional dynamic feature matrix including a time axis, etc.; determines the initial operation constraint conditions according to the valve body life distribution model and the switching mode library; generates an initial energy consumption curve through dynamic switching simulation; and uses an adaptive optimization algorithm to iteratively correct and generate a multi-dimensional energy-saving operation strategy. The system includes modules such as data processing, matrix construction, condition determination, simulation, and optimization, and cooperates to implement the above method. The present invention can accurately analyze the operation state of the valve body, optimize the operation strategy, effectively reduce energy consumption, improve the operation performance and the equipment management level, is applicable to various industrial scenarios, and has good application prospects.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent valve body control, and specifically provides an energy-saving intelligent switching valve body operation method and system. Background Art

[0002] In modern industrial production, as a key component of the fluid transportation system, intelligent switching valve bodies are widely used in many fields such as petrochemical, power, and metallurgy. The quality of their performance directly affects the stability, safety, and energy utilization efficiency of the production process.

[0003] With the continuous expansion of industrial production scale and the increasing complexity of production processes, the operating requirements for intelligent switching valve bodies are also getting higher and higher. Most traditional valve body operation methods rely on manual experience or simple control logic, and it is difficult to adapt to complex and changeable working conditions. On the one hand, it is impossible to accurately conduct comprehensive regulation based on multiple factors such as fluid pressure fluctuations, spool wear, medium temperature changes, and actuator energy consumption. As a result, during actual operation, problems such as frequent leakage and slow response of the valve body often occur, which not only affects production efficiency but also increases equipment maintenance costs and safety risks. For example, in petrochemical production, if the valve body cannot be switched in a timely and accurate manner, it may lead to an imbalance in the fluid pressure in the pipeline, triggering serious accidents such as material leakage and even explosion.

[0004] On the other hand, the energy issue has become the focus of global attention. As the main field of energy consumption, energy conservation and emission reduction in industrial production are extremely urgent. However, there are many unreasonable aspects in the existing valve body operation methods in terms of energy utilization. Due to the lack of effective monitoring and optimization of actuator energy consumption, and the failure to fully consider the impact of factors such as spool wear on energy consumption, the valve body consumes a large amount of unnecessary energy during operation. Taking the power industry as an example, a large number of intelligent switching valve bodies cause huge energy waste due to excessive energy consumption during long-term operation, increasing the power generation cost, which runs counter to the concept of sustainable development.

[0005] In addition, with the development trend of industrial automation and intelligence, the existing valve body operation methods and systems have been difficult to meet the needs of intelligent production. It is impossible to achieve in-depth mining and analysis of valve body operation data, and it cannot provide strong support for production decision-making. In the information age, enterprises need real-time and accurate data to optimize production processes and improve production management levels. However, traditional valve body operation methods cannot provide these key data, restricting the intelligent upgrading of enterprises. Summary of the Invention

[0006] The purpose of the present invention is to provide an energy-saving intelligent switching valve body operation method and system to solve the problems raised in the above background art.

[0007] To achieve the above object, the present invention provides the following technical solutions: An energy-saving intelligent switching valve body operation method, the method comprising:

[0008] Obtain the historical operation data of the intelligent switching valve body, and classify the data according to preset dimensions to generate a structured feature set; the preset dimensions include a fluid pressure fluctuation curve, a spool wear degree, a medium temperature change, and an actuator energy consumption spectrum;

[0009] Construct a dynamic feature matrix, and define a multi-dimensional analysis coordinate system based on the classification dimensions of the feature set. The coordinate system includes a time axis, a fluid axis, a mechanical axis, and an energy consumption axis;

[0010] Determine the initial operation constraint conditions according to the valve body life distribution model and the switching mode library. The constraint conditions include a leakage probability threshold, a switching response time rule, and an energy consumption sensitivity classification;

[0011] Perform dynamic switching simulation on the feature set based on the dynamic feature matrix, simulate the switching behavior of the valve body under various working conditions through a multi-modal simulation engine, and generate an initial energy consumption curve;

[0012] Adopt an adaptive optimization algorithm to iteratively correct the initial energy consumption curve, adjust the spool wear accumulation weight and the energy consumption coupling coefficient, and generate a multi-dimensional energy-saving operation strategy.

[0013] Preferably, the constructing of the dynamic feature matrix includes:

[0014] Divide the time axis into switching stages synchronized with the medium transportation period, and each stage is associated with a pressure transient compensation factor;

[0015] Define a time-varying response function of the spool displacement based on the fluid axis, and integrate the medium temperature lag characteristic;

[0016] Embed the energy consumption spectrum density distribution of the actuator in the mechanical axis, and dynamically associate the switching times with the mechanical efficiency correction coefficient.

[0017] Preferably, the adaptive optimization algorithm adopts an improved random forest-long short-term memory hybrid model, including:

[0018] Encode the spool displacement deviation and the actuator energy consumption decay rate as a multi-objective optimization function, and define a loss function to evaluate the energy consumption sensitivity matching degree and the mechanical parameter drift cost;

[0019] Extract high-dimensional feature interaction terms through a gradient boosting decision tree, and adopt a time series memory unit to capture dynamic dependence relationships, and output an energy-saving prediction value integrating the energy consumption trend.

[0020] Preferably, the division of the time axis includes:

[0021] Based on the periodic characteristics of historical operation data, the time axis is divided into dynamic sliding windows, and each window is associated with the start and stop marks of the medium transportation stage;

[0022] A time decay factor is introduced to dynamically adjust the weight distribution of each stage, and the hidden Markov model is used to predict the switching requirements of future time segments.

[0023] Preferably, the definition of the fluid axis includes:

[0024] A non-linear regression constraint equation for spool displacement is defined according to the change rate of medium viscosity, and the flow velocity distribution function is optimized based on fluid-structure interaction analysis;

[0025] The fluid pulsation energy spectrum is extracted through the frequency-domain integration algorithm, and the response delay parameter of the medium temperature lag characteristic is dynamically corrected.

[0026] Preferably, the definition of the mechanical axis includes:

[0027] The wavelet packet energy entropy algorithm is used to quantify the high-frequency components of the vibration signal of the actuator, and the vibration energy is correlated with the mechanical wear rate in real time;

[0028] The attenuation gradient of the mechanical efficiency correction coefficient is dynamically adjusted according to the stress distribution model of the sealing component.

[0029] Preferably, the definition of the energy consumption axis includes:

[0030] Based on the spatio-temporal distribution characteristics of historical energy consumption data, a hierarchical clustering model of the energy consumption benchmark curve is established;

[0031] The energy consumption abnormal events are identified through the mutation detection mechanism, and the adaptive threshold segmentation algorithm is used to generate the dynamic interval division of the energy consumption sensitivity.

[0032] Preferably, the division of the dynamic sliding window further includes:

[0033] The window length is adaptively adjusted according to the pipeline pressure fluctuation amplitude, and the gradient descent method is introduced to optimize the iteration step size of the time decay factor;

[0034] The window boundary is aligned with the valve body action timing sequence to generate an energy consumption evaluation matrix with joint calibration of the time axis and the mechanical axis.

[0035] Preferably, the iterative correction includes:

[0036] A dual-stream attention network model is constructed, with the spool friction energy, the current distortion rate of the actuator, and the medium corrosion index as input features;

[0037] The key energy consumption factors are screened through the adaptive feature selection mechanism, and the multi-head attention layer is used to fuse the cross-axis domain correlation features to generate the corrected energy consumption confidence interval.

[0038] Preferably, the present invention further includes an energy-saving intelligent switching valve body operating system, and the system includes:

[0039] A data processing module, configured to obtain the historical operation data of the intelligent switching valve body, and generate a structured feature set according to the preset dimensions of the fluid pressure fluctuation curve, the spool wear degree, the medium temperature change, and the actuator energy consumption spectrum;

[0040] A matrix construction module, configured to define a multi-dimensional analysis coordinate system including a time axis, a fluid axis, a mechanical axis, and an energy consumption axis based on the classification dimensions of the feature set, and construct a dynamic feature matrix;

[0041] A condition determination module, configured to determine initial operation constraint conditions including a leakage probability threshold, a switching response time rule, and an energy consumption sensitivity classification according to the valve body life distribution model and the switching mode library;

[0042] A simulation module, configured to perform dynamic switching simulation on the feature set based on the dynamic feature matrix, simulate the switching behavior of the valve body under various working conditions through a multi-modal simulation engine, and generate an initial energy consumption curve;

[0043] An optimization module, configured to iteratively correct the initial energy consumption curve by using an adaptive optimization algorithm, adjust the spool wear accumulation weight and the energy consumption coupling coefficient, and generate a multi-dimensional energy-saving operation strategy.

[0044] Compared with the prior art, the beneficial effects of the present invention are:

[0045] At the level of data utilization and analysis, by obtaining the historical operation data of the intelligent switching valve body and generating a structured feature set according to the preset dimensions such as the fluid pressure fluctuation curve, the spool wear degree, the medium temperature change, and the actuator energy consumption spectrum, the comprehensive collection and in-depth collation of the valve body operation information are realized. This lays a solid foundation for accurately analyzing the valve body operation state subsequently. Compared with the traditional method that only relies on single or a small number of parameters for judgment, the accuracy and comprehensiveness of the valve body operation condition assessment are greatly improved. Based on these structured feature sets, a dynamic feature matrix is constructed, and using the multi-dimensional analysis coordinate system, the changing trends of different operation parameters over time and their mutual relationships can be clearly displayed, enabling operators to more intuitively and deeply understand the operation rules of the valve body, thereby providing strong support for formulating reasonable operation strategies.

[0046] Regarding the valve body life and operation reliability, initial operation constraint conditions are determined according to the valve body life distribution model and the switching mode library, including the leakage probability threshold, switching response time rule, energy consumption sensitivity classification, etc. This measure effectively avoids the problems of premature damage to the valve body or increased leakage risk caused by improper operation. By strictly controlling the leakage probability threshold, it ensures that the valve body operates in a safe and reliable state, reducing material losses and environmental pollution caused by leakage. At the same time, the reasonable switching response time rule ensures that the valve body can act quickly and accurately when switching is required, improving the stability and continuity of the production process. For example, in some pharmaceutical production lines with extremely high requirements for flow control accuracy, the present invention can ensure that the valve body responds promptly to the flow change instruction, guaranteeing the quality stability of drug production. The setting of the energy consumption sensitivity classification provides a clear direction and basis for formulating energy-saving operation strategies.

[0047] In terms of energy-saving optimization, dynamic switching simulation is performed based on the dynamic feature matrix to generate the initial energy consumption curve, and an adaptive optimization algorithm is used to iteratively correct it. This method can fully consider factors such as the cumulative weight of spool wear and the energy consumption coupling coefficient. By adjusting these parameters, the operation strategy of the valve body is optimized, achieving significant energy-saving effects. Taking the fluid transportation system of a large steel enterprise as an example, after applying the technology of the present invention, a large amount of energy costs can be saved every year, reducing the production costs of the enterprise and improving the market competitiveness of the enterprise. In addition, by identifying energy consumption abnormal events and dividing the dynamic interval of energy consumption sensitivity, the problem of excessive energy consumption can be discovered and solved in a timely manner, further tapping the energy-saving potential.

[0048] From the perspective of improving the overall operation performance, improved models in the adaptive optimization algorithm, such as encoding the spool displacement deviation and the energy consumption decay rate of the actuator as a multi-objective optimization function, and using gradient boosting decision trees and temporal memory units, can accurately capture the dynamic changes and complex relationships during the operation of the valve body. By screening key energy consumption factors through the adaptive feature selection mechanism and fusing cross-axis domain correlation features, not only a more accurate energy consumption confidence interval is generated, but also a multi-dimensional energy-saving operation strategy is formulated. This enables the valve body to operate efficiently under different working conditions, extends the service life of the valve body, reduces the frequency of equipment maintenance and replacement, improves the operation efficiency and reliability of the entire fluid transportation system, and provides a strong guarantee for the stable and efficient operation of industrial production. Brief Description of the Drawings

[0049] Figure 1 It is the working principle diagram of the energy-saving intelligent switching valve body operation method described in the present invention;

[0050] Figure 2 It is the flowchart of the adaptive optimization algorithm;

[0051] Figure 3Flow chart for time axis division;

[0052] Figure 4 Flow chart defined for the energy consumption axis. Specific implementation manners

[0053] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0054] Please refer to Figures 1-4 , the present invention relates to an energy-saving intelligent switching valve body operation method, and its specific implementation manners are elaborated in detail below.

[0055] Obtain the historical operation data of the intelligent switching valve body, and classify the data according to preset dimensions to generate a structured feature set. Among them, the preset dimensions include the fluid pressure fluctuation curve, the spool wear degree, the medium temperature change, and the actuator energy consumption spectrum. These data are the basis for subsequent analysis and decision-making. By collecting and organizing the historical operation data, the operation characteristics of the valve body under different working conditions can be deeply understood. For example, obtain the fluid pressure fluctuation conditions in different time periods from the long-term operation records, accurately record the wear degree of the spool in each stage, monitor the change trend of the medium temperature over time, and the energy consumption of the actuator under different operations.

[0056] Construct a dynamic feature matrix, and define a multi-dimensional analysis coordinate system based on the classification dimensions of the feature set. This coordinate system includes a time axis, a fluid axis, a mechanical axis, and an energy consumption axis. The multi-dimensional analysis coordinate system provides a powerful tool for comprehensively analyzing the operation state of the valve body. The time axis is used to record the time sequence of the valve body operation, the fluid axis reflects the parameter changes related to the fluid, the mechanical axis reflects the operation characteristics of the mechanical components, and the energy consumption axis focuses on the analysis of the energy consumption data. By integrating the data of different dimensions into a matrix, the mutual relationship between various factors can be observed more intuitively, providing comprehensive data support for subsequent analysis and decision-making.

[0057] Determine the initial operation constraint conditions according to the valve body life distribution model and the switching mode library. The constraint conditions include the leakage probability threshold, the switching response time rule, and the energy consumption sensitivity classification. The valve body life distribution model is established based on a large amount of experimental data and theoretical analysis, and is used to predict the life of the valve body under different usage conditions. The switching mode library stores a variety of verified switching modes. The initial operation constraint conditions determined based on this information can ensure the safe and reliable operation of the valve body, while meeting the requirements of energy conservation and high efficiency. For example, setting a reasonable leakage probability threshold to ensure that the valve body operates within a certain leakage risk range; formulating a switching response time rule to ensure that the valve body can respond in a timely manner when switching is required; classifying the energy consumption sensitivity to more targeted reduce energy consumption when optimizing the operation strategy.

[0058] Perform dynamic switching simulation on the feature set based on the dynamic feature matrix. Simulate the switching behavior of the valve body under various working conditions through a multi-modal simulation engine to generate the initial energy consumption curve. The multi-modal simulation engine can comprehensively consider various factors, such as the flow characteristics of the fluid, the mechanical properties of mechanical components, and the change law of energy consumption, to simulate the actual switching process of the valve body under different working conditions. Through a large number of simulation experiments, the initial energy consumption curves under different working conditions can be obtained. These curves reflect the energy consumption of the valve body under the current operating conditions and provide a reference basis for subsequent optimization.

[0059] Use an adaptive optimization algorithm to iteratively correct the initial energy consumption curve, adjust the cumulative weight of spool wear and the energy consumption coupling coefficient, and generate a multi-dimensional energy-saving operation strategy. The adaptive optimization algorithm can continuously adjust the optimization parameters according to the actual situation to achieve the best energy-saving effect. By adjusting the cumulative weight of spool wear and the energy consumption coupling coefficient, the operation strategy of the valve body can be optimized, unnecessary energy consumption can be reduced, and the service life of the valve body can be extended at the same time. For example, dynamically adjust the weight of the spool in the energy consumption calculation according to the wear degree of the spool to make the operation strategy more in line with the actual situation; optimize the energy consumption coupling coefficient to improve the energy utilization efficiency and achieve multi-dimensional energy-saving goals.

[0060] The following further details the specific implementation manners of the present invention through 6 embodiments.

[0061] Example 1: In the process of constructing the dynamic feature matrix, the division of the time axis is involved. Specifically, based on the periodic characteristics of historical operation data, the time axis is divided into dynamic sliding windows. This is because the operation of the valve body often has a certain periodicity. For example, in industrial production, the transportation of the medium may follow a certain time pattern. By analyzing the historical operation data, this periodic characteristic can be determined, and then the time axis can be divided into switching stages synchronized with the medium transportation cycle. Each window is associated with the start and stop marks of the medium transportation stage, so that the actual production stage corresponding to each window can be clarified, which is convenient for subsequent analysis. For example, in a chemical production process, the medium transportation is divided into stages such as feeding, reaction, and discharging, and the time and operation requirements of each stage are different. By marking the start and stop times, the operation of the valve body in different stages can be accurately understood.

[0062] Introduce a time decay factor to dynamically adjust the weight distribution of each stage. As time goes by, the influence of data in earlier stages on the current decision may gradually decrease, so it is necessary to introduce a time decay factor. The value range of the time decay factor is usually between 0 and 1, and its calculation formula is: , where represents the time decay factor, is the decay coefficient, which is determined according to the actual situation, represents the time interval between the current time and the data occurrence time. By adjusting the value, the decay speed can be controlled. Use the hidden Markov model to predict the switching requirements of future time segments. The hidden Markov model is a statistical model that can predict future states based on historical data. In this example, through the training of historical operation data, the hidden Markov model can learn the switching rules of the valve body, so as to predict possible future switching requirements. For example, according to the switching frequency and time interval of the valve body under different working conditions in the past period of time, the model can predict whether the valve body needs to be switched in the next few hours and at what time point it is most appropriate to switch. In this way, preparations can be made in advance, the operation of the valve body can be optimized, and energy consumption can be reduced.

[0063] In addition, the window length is adaptively adjusted according to the amplitude of the pipeline pressure fluctuation. When the amplitude of the pipeline pressure fluctuation is large, it indicates that the operating conditions change violently, and it is necessary to observe the operating state of the valve body more finely. Therefore, the window length is appropriately shortened; conversely, when the amplitude of the pressure fluctuation is small, the window length can be appropriately extended. The gradient descent method is introduced to optimize the iteration step size of the time decay factor. The gradient descent method is a commonly used optimization algorithm that gradually reduces the objective function by continuously adjusting the values of the parameters. In this embodiment, the objective function can be a function related to energy consumption. By adjusting the iteration step size of the time decay factor, the time decay factor can converge to the optimal value faster, so as to more accurately adjust the weight distribution of each stage. Finally, the window boundary is aligned with the valve body action time sequence to generate an energy consumption evaluation matrix with joint calibration of the time axis and the mechanical axis. This can ensure that when evaluating the energy consumption, the impacts of time and mechanical factors can be comprehensively considered, and the energy consumption of the valve body can be analyzed more accurately.

[0064] Embodiment 2: For the definition of the fluid axis, a non-linear regression constraint equation for the spool displacement is defined according to the change rate of the medium viscosity. The medium viscosity is one of the important factors affecting fluid flow and spool operation. In actual operation, the medium viscosity may change with factors such as temperature and pressure. Let the change rate of the medium viscosity be , and the spool displacement be . Through a large number of experiments and data analysis, the following non-linear regression constraint equation can be established: , where is the coefficient fitted according to the experimental data. This equation reflects the relationship between the change rate of the medium viscosity and the spool displacement. In actual operation, the reasonable spool displacement can be calculated through this equation according to the real-time monitored change rate of the medium viscosity to ensure the normal operation of the valve body.

[0065] The flow velocity distribution function is optimized based on the fluid-structure interaction analysis. The fluid-structure interaction analysis considers the interaction between the fluid and the solid. During the operation of the valve body, the fluid flow will exert forces on solid components such as the spool, and at the same time, the movement of the solid components will also affect the fluid flow state. Through the fluid-structure interaction analysis, a more accurate flow velocity distribution can be obtained, so as to optimize the flow velocity distribution function. For example, the finite element analysis method is used to simulate the flow field inside the valve body, considering factors such as the shape and position of the spool and the physical properties of the fluid, to obtain the flow velocity distribution data under different operating conditions. According to these data, a flow velocity distribution function is established, where represents the flow velocity, represents the spatial coordinates, and represents the time. By optimizing this function, the fluid flow efficiency can be improved and the energy consumption can be reduced.

[0066] Extract the fluid pulsation energy spectrum through the frequency-domain integration algorithm and dynamically correct the response delay parameter of the medium temperature lag characteristic. Fluid pulsation will affect the operation of the valve body. Through the frequency-domain integration algorithm, the fluid pulsation signal in the time domain can be converted to the frequency domain to extract its energy spectrum. Let the fluid pulsation signal be , and its frequency spectrum is obtained through Fourier transform , then the fluid pulsation energy spectrum can be calculated through frequency-domain integration: , where and are the frequency ranges of integration. According to the extracted energy spectrum, the frequency and energy distribution of fluid pulsation can be analyzed. The medium temperature lag characteristic means that the change of the medium temperature has a certain delay relative to the change of the external conditions. By analyzing the fluid pulsation energy spectrum, the response delay parameter of the medium temperature lag characteristic can be dynamically corrected, making the monitoring and control of the medium temperature change more accurate, and then optimizing the operation strategy of the valve body.

[0067] Example 3: In terms of the definition of the mechanical shaft, the wavelet packet energy entropy algorithm is used to quantify the high-frequency components of the vibration signal of the actuator. The actuator will generate vibration during operation. The vibration signal contains rich information, reflecting the operating state of the actuator. The wavelet packet energy entropy algorithm is an effective signal analysis method. It can perform multi-resolution decomposition on the signal and extract the energy information of different frequency bands. Let the vibration signal of the actuator be , after wavelet packet decomposition, the signal components of different frequency bands are obtained , . The energy of each frequency band can be calculated as: , where and are the time ranges of integration. Then, calculate the wavelet packet energy entropy . The larger the wavelet packet energy entropy , the higher the uncertainty of the signal and the richer the high-frequency components. By real-time monitoring the change of the wavelet packet energy entropy , the vibration energy can be real-time correlated with the mechanical wear rate. When value increases, it means that the vibration energy of the actuator increases, which may lead to an accelerated mechanical wear rate. At this time, corresponding measures need to be taken, such as adjusting the operation parameters or performing equipment maintenance.

[0068] Dynamically adjust the attenuation gradient of the mechanical efficiency correction coefficient according to the stress distribution model of the sealing component. The sealing component is an important part of the valve body, and its stress distribution will affect the mechanical efficiency and sealing performance. By establishing the stress distribution model of the sealing component, the stress distribution of the sealing component under different working conditions can be analyzed. Let the stress of the sealing component under a certain working condition be Based on the stress distribution model, the relationship between stress and the mechanical efficiency correction factor is obtained. When the stress exceeds a certain threshold, the mechanical efficiency correction factor will decay, and its decay gradient can be dynamically adjusted according to the stress distribution model. For example, when the stress is relatively large, the decay gradient is appropriately increased so that the mechanical efficiency correction factor decays faster to reflect the actual decrease in mechanical efficiency. In this way, when calculating and optimizing the operation strategy of the valve body, the influence of mechanical factors can be more accurately considered, improving the energy-saving effect and the service life of the valve body.

[0069] Example 4: For the definition of the energy consumption axis, based on the spatio-temporal distribution characteristics of historical energy consumption data, a hierarchical clustering model of the energy consumption benchmark curve is established. The historical energy consumption data contains the energy consumption information of the valve body under different time and space conditions. The analysis of spatio-temporal distribution characteristics can help understand the variation law of energy consumption, such as the energy consumption differences in different seasons, different production periods, and different working areas. The hierarchical clustering algorithm is used to process these data, and this algorithm can divide the data into different hierarchical structures according to similarity. Let the historical energy consumption data be , where represents the th energy consumption data point. First, each data point is regarded as a separate class, and then the distance between classes is calculated. The calculation method of the distance can adopt common distance metrics such as Euclidean distance. For example, the Euclidean distance between two data points and . According to the distance, the two closest classes are merged into a new class, and this process is continuously repeated until all data points are merged into a large class, thus obtaining the hierarchical clustering result of the energy consumption benchmark curve. Through this model, the energy consumption benchmark level under different working conditions can be determined, providing a reference for subsequent energy consumption analysis and optimization.

[0070] Identify energy consumption abnormal events through a mutation detection mechanism, and use an adaptive threshold segmentation algorithm to generate a dynamic interval division of energy consumption sensitivity. The mutation detection mechanism is used to monitor sudden changes in energy consumption data. When there are abnormal fluctuations in energy consumption data, it may mean that there are problems in the operation of the valve body or the working conditions have changed significantly. For example, the cumulative sum control chart (CUSUM) method is used for mutation detection. Let the energy consumption data sequence be , and define the cumulative sum statistic , where is the average energy consumption under normal conditions. When When it exceeds a certain threshold, it is determined that an energy consumption mutation event has occurred. For the identified abnormal energy consumption events, an adaptive threshold segmentation algorithm is used to generate a dynamic interval division of energy consumption sensitivity. The adaptive threshold segmentation algorithm automatically adjusts the threshold according to the data distribution, such as using the Otsu algorithm. This algorithm determines the optimal threshold by maximizing the between-class variance. Let the gray level of the energy consumption data be , the number of pixels for each gray level be , and the total number of pixels be . Define the probability of the gray level . Assume that the gray levels are divided into two classes and , the threshold is , then the between-class variance , where , . By calculating the between-class variances under different thresholds, find the threshold that maximizes the between-class variance, and use this to divide the dynamic interval of energy consumption sensitivity. In this way, according to the changes in energy consumption, the energy consumption sensitivity can be evaluated more accurately, providing a basis for formulating energy-saving operation strategies.

[0071] Example 5: When using the adaptive optimization algorithm to iteratively correct the initial energy consumption curve, a dual-stream attention network model is constructed. This model uses the spool friction energy, the actuator current distortion rate, and the medium corrosion index as input features. The spool friction energy reflects the energy loss generated by the spool due to friction during movement, the actuator current distortion rate reflects whether the working state of the actuator is normal, and the medium corrosion index indicates the corrosion degree of the medium on the valve body components. These features are all closely related to the energy consumption of the valve body.

[0072] Key energy consumption factors are screened through an adaptive feature selection mechanism. The adaptive feature selection mechanism can automatically select features that have a greater impact on energy consumption according to the characteristics of the data and the requirements of the model. For example, the feature importance evaluation method in the random forest algorithm is used. Random forest is an ensemble learning model composed of multiple decision trees. By calculating the importance scores of each feature in the decision trees, the importance of the features is evaluated. Let the importance score of the feature , and the calculation method is that in all decision trees, the feature The sum of the contributions to node splitting. Select features with higher importance scores as key energy consumption factors. Use a multi-head attention layer to fuse cross-axis domain correlation features and generate a corrected energy consumption confidence interval. The multi-head attention layer can simultaneously focus on the relationships between different features and fuse cross-axis domain correlation features from multiple perspectives. For example, among the time axis, fluid axis, mechanical axis, and energy consumption axis, there may be mutual influences between the features on different axes. Through the multi-head attention layer, these cross-axis domain correlation features can be fused to obtain more comprehensive information. Let the input of the multi-head attention layer be and the output be . Its calculation formula is: , where is the number of heads, , and are learnable weight matrices, and is the dimension of the key vector. Through the fusion of the multi-head attention layer, more accurate energy consumption information can be obtained, and then a corrected energy consumption confidence interval can be generated to support the formulation of more precise energy-saving operation strategies.

[0073] Embodiment 6: The present invention also relates to an energy-saving intelligent switching valve body operating system. The data processing module of this system is used to obtain the historical operation data of the intelligent switching valve body and generate a structured feature set by classifying according to the preset dimensions of the fluid pressure fluctuation curve, spool wear degree, medium temperature change, and actuator energy consumption spectrum. In practical applications, the data processing module can be connected to various sensors and data acquisition devices to collect various data during the operation of the valve body in real time. For example, collect fluid pressure fluctuation data through a pressure sensor, obtain spool wear degree data through a wear detection device, monitor medium temperature change data through a temperature sensor, and record actuator energy consumption spectrum data through an energy consumption monitoring device. Then, clean, organize, and classify these original data and convert them into a structured feature set for subsequent analysis and processing.

[0074] The matrix construction module defines a multi-dimensional analysis coordinate system including a time axis, a fluid axis, a mechanical axis, and an energy consumption axis based on the classification dimensions of the feature set and constructs a dynamic feature matrix. The matrix construction module constructs a dynamic feature matrix according to the structured feature set provided by the data processing module and in combination with the definition methods of the time axis, fluid axis, mechanical axis, and energy consumption axis in the present invention. For example, when constructing the time axis, divide it according to the method in Embodiment 1; when constructing the fluid axis, define it according to the method in Embodiment 2; the construction of the mechanical axis and the energy consumption axis respectively refer to the relevant content of Embodiment 3 and Embodiment 4. By constructing a dynamic feature matrix, information from different dimensions is integrated together, providing a unified data platform for subsequent analysis and decision-making.

[0075] The condition determination module determines the initial operation constraint conditions including the leakage probability threshold, switching response time rule, and energy consumption sensitivity classification based on the valve body life distribution model and the switching mode library. The valve body life distribution model and the switching mode library can be pre-stored in the system database. The condition determination module retrieves the corresponding model and library information from the database according to the current operation state and historical data of the valve body, and calculates and determines the initial operation constraint conditions. For example, according to factors such as the service life and operation environment of the valve body, combined with the valve body life distribution model, a reasonable leakage probability threshold is determined; according to the production process requirements and past experience, a suitable switching mode is selected from the switching mode library, and the corresponding switching response time rule is formulated; according to the analysis results of energy consumption data and the standards of energy consumption sensitivity classification, the current energy consumption sensitivity classification is determined. These initial operation constraint conditions provide an important basis for subsequent simulation and optimization.

[0076] The simulation module performs dynamic switching simulation on the feature set based on the dynamic feature matrix, and simulates the switching behavior of the valve body under various working conditions through a multi-modal simulation engine to generate the initial energy consumption curve. The simulation module uses the dynamic feature matrix generated by the matrix construction module and the feature set provided by the data processing module, and performs simulation with the help of the multi-modal simulation engine. The multi-modal simulation engine integrates a variety of physical models and algorithms, and can comprehensively consider various factors such as fluid mechanics, mechanical dynamics, and thermodynamics. For example, during the simulation of the valve body switching process, the forces acting on the valve core due to changes in fluid pressure, flow rate, and viscosity are considered, as well as the influence of the valve core movement on the fluid flow state; at the same time, the mechanical and electrical characteristics of the actuator and their interactions with the mechanical components of the valve body are considered. By conducting a large number of simulation experiments under different working conditions, such as different medium flow rates, pressures, temperatures, different initial positions and movement speeds of the valve core, etc., the corresponding initial energy consumption curves are generated. These curves intuitively show the energy consumption changes during the valve body switching process under the current operation conditions, providing basic data for subsequent optimization.

[0077] The optimization module uses an adaptive optimization algorithm to iteratively correct the initial energy consumption curve, adjust the cumulative weight of spool wear and the energy consumption coupling coefficient, and generate a multi-dimensional energy-saving operation strategy. The optimization module optimizes according to the initial energy consumption curve generated by the simulation module using an adaptive optimization algorithm. For example, the improved random forest-long short-term memory hybrid model mentioned in Embodiment 5 is used as the adaptive optimization algorithm. This algorithm encodes the spool displacement deviation and the energy consumption decay rate of the actuator into a multi-objective optimization function and defines a loss function to evaluate the matching degree of energy consumption sensitivity and the mechanical parameter drift cost. High-dimensional feature interaction terms are extracted through gradient boosting decision trees, and temporal memory units are used to capture dynamic dependencies, and an energy-saving prediction value integrating the energy consumption trend is output. During the iteration process, the cumulative weight of spool wear and the energy consumption coupling coefficient are continuously adjusted to find the optimal operation strategy. After each iteration, a new simulation is performed according to the new operation strategy to obtain a new energy consumption curve, which is compared with the previous result. If the energy consumption is reduced or other optimization goals are met, the next iteration continues; otherwise, the parameters of the optimization algorithm are adjusted or other optimization strategies are adopted until the optimal multi-dimensional energy-saving operation strategy is found. The finally generated energy-saving operation strategy covers multiple dimensions such as time, fluid, machinery, and energy consumption, and can minimize energy consumption and improve the overall performance and economic benefits of the system while ensuring the normal operation of the valve body.

[0078] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements but also other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0079] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An energy-saving intelligent switching valve body operation method, characterized in that, Including: Obtain the historical operation data of the intelligent switching valve body, classify the data according to preset dimensions to generate a structured feature set; The preset dimensions include fluid pressure fluctuation curve, spool wear degree, medium temperature change, and actuator energy consumption spectrum; Construct a dynamic feature matrix, define a multi-dimensional analysis coordinate system based on the classification dimensions of the feature set, and the coordinate system includes a time axis, a fluid axis, a mechanical axis, and an energy consumption axis; Determine the initial operation constraint conditions according to the valve body life distribution model and the switching mode library, and the constraint conditions include leakage probability threshold, switching response time rule, and energy consumption sensitivity classification; Perform dynamic switching simulation on the feature set based on the dynamic feature matrix, simulate the switching behavior of the valve body under various working conditions through a multi-modal simulation engine, and generate an initial energy consumption curve; Adopt an adaptive optimization algorithm to iteratively correct the initial energy consumption curve, adjust the spool wear accumulation weight and the energy consumption coupling coefficient, and generate a multi-dimensional energy-saving operation strategy.

2. The energy-saving intelligent switching valve body operation method according to claim 1, characterized in that, The construction of the dynamic feature matrix includes: Divide the time axis into switching stages synchronized with the medium transportation period, and each stage is associated with a pressure transient compensation factor; Define a time-varying response function of the spool displacement based on the fluid axis, and integrate the medium temperature lag characteristic; Embed the energy consumption spectrum density distribution of the actuator in the mechanical axis, and dynamically associate the switching times with the mechanical efficiency correction coefficient.

3. The energy-saving intelligent switching valve body operation method according to claim 1, characterized in that, The adaptive optimization algorithm adopts an improved random forest-long short-term memory hybrid model, including: Encode the spool displacement deviation and the actuator energy consumption decay rate as a multi-objective optimization function, and define a loss function to evaluate the energy consumption sensitivity matching degree and the mechanical parameter drift cost; Extract high-dimensional feature interaction terms through gradient boosting decision trees, and use a time series memory unit to capture dynamic dependencies, and output an energy-saving prediction value that fuses the energy consumption trend.

4. The energy-saving intelligent switching valve body operation method according to claim 2, characterized in that, The division of the time axis includes: Based on the periodic characteristics of the historical operation data, divide the time axis into dynamic sliding windows, and each window is associated with the start and stop marks of the medium transportation stage; Introduce a time decay factor to dynamically adjust the weight distribution of each stage, and use a hidden Markov model to predict the switching demand of future time segments.

5. The energy-saving intelligent switching valve body operation method according to claim 2, characterized in that, The definition of the fluid axis includes: Define a non-linear regression constraint equation for the spool displacement according to the medium viscosity change rate, and optimize the flow velocity distribution function based on the fluid-structure interaction analysis; Extract the fluid pulsation energy spectrum through the frequency domain integration algorithm, and dynamically correct the response delay parameter of the medium temperature lag characteristic.

6. The energy-saving intelligent switching valve body operation method according to claim 1, wherein, The definition of the mechanical axis includes: Use the wavelet packet energy entropy algorithm to quantify the high-frequency components of the actuator vibration signal, and real-time associate the vibration energy with the mechanical wear rate; Dynamically adjust the attenuation gradient of the mechanical efficiency correction coefficient according to the seal component stress distribution model.

7. The energy-saving intelligent switching valve body operation method according to claim 2, characterized in that The definition of the energy consumption axis includes: Based on the spatio-temporal distribution characteristics of the historical energy consumption data, establish a hierarchical clustering model of the energy consumption benchmark curve; Identify energy consumption abnormal events through a mutation detection mechanism, and use an adaptive threshold segmentation algorithm to generate a dynamic interval division of the energy consumption sensitivity.

8. The energy-saving intelligent switching valve body operation method according to claim 4, characterized in that The division of the dynamic sliding window also includes: Adaptive adjust the window length according to the pipeline pressure fluctuation amplitude, and introduce the gradient descent method to optimize the iteration step of the time decay factor; Align the window boundary with the valve body action timing sequence to generate an energy consumption evaluation matrix for time-axis and mechanical-axis joint calibration.

9. The energy-saving intelligent switching valve body operation method according to claim 1, characterized in that The iterative correction includes: Construct a dual-stream attention network model with the spool friction energy, actuator current distortion rate, and medium corrosion index as input features; Screen key energy consumption factors through an adaptive feature selection mechanism, and use a multi-head attention layer to fuse cross-axis domain correlation features to generate a corrected energy consumption confidence interval.

10. An energy-saving intelligent switching valve body operating system, characterized in that, It includes: A data processing module for obtaining the historical operation data of the intelligent switching valve body and generating a structured feature set classified according to the preset dimensions of the fluid pressure fluctuation curve, spool wear degree, medium temperature change, and actuator energy consumption spectrum; A matrix construction module for defining a multi-dimensional analysis coordinate system including a time axis, a fluid axis, a mechanical axis, and an energy consumption axis based on the classification dimensions of the feature set, and constructing a dynamic feature matrix; A condition determination module for determining initial operation constraint conditions including a leakage probability threshold, a switching response time rule, and an energy consumption sensitivity classification according to the valve body life distribution model and the switching mode library; A simulation module for performing dynamic switching simulation on the feature set based on the dynamic feature matrix, simulating the switching behavior of the valve body under various working conditions through a multi-modal simulation engine, and generating an initial energy consumption curve; An optimization module for iteratively correcting the initial energy consumption curve using an adaptive optimization algorithm, adjusting the spool wear accumulation weight and the energy consumption coupling coefficient, and generating a multi-dimensional energy-saving operation strategy.

Images