Precise adjustment system and method of electric butterfly valve based on pipeline multivariable cooperation

Abstract

The present application relates to a kind of based on pipeline multivariate coordination electric butterfly valve precision regulating system and method, belong to industrial automatic control technical field.The system includes: multivariate perception unit gathers pipeline multivariate parameter and carries out noise reduction, coupling correlation analysis and normalization, generates multivariate coordination feature data;Cooperative decision unit constructs multi-scenario condition benchmark database, generates the cooperative control parameter of butterfly valve opening and rate by condition similarity matching and deviation weight dynamic distribution;Butterfly valve execution unit carries out self-adapting zero and stroke calibration, and based on cooperative control parameter combination real-time feedback to valve plate carries out gradient regulation and dynamic correction;Closed loop optimization unit carries out coupling deviation separation and quantization check to the parameter after regulation, forms closed loop iterative optimization.Electric butterfly valve is realized in multivariate coupling and variable condition condition whole life cycle precision, adaptive control, improves the regulating precision and robustness of pipeline system.

Description

Technical Field

[0001] This invention belongs to the field of industrial automatic control technology, specifically relating to a precise adjustment system and method for electric butterfly valves based on multi-variable pipeline coordination. Background Technology

[0002] Electric butterfly valves, as key actuators in industrial automated pipeline systems, are widely used for regulating the flow and pressure of media in fields such as petroleum, chemical, power, metallurgy, and municipal water supply and drainage. By controlling the rotation angle of the valve plate to change the flow area of ​​the pipeline, they achieve continuous control of fluid conditions, which is of great significance for ensuring process stability and improving energy efficiency. With the increasing demands of modern industry for control precision, response speed, and adaptability, the control performance of electric butterfly valves directly determines the operational quality of the entire pipeline system.

[0003] Currently, conventional electric butterfly valves mainly employ two control methods: on / off and regulating. On / off actuators can only achieve full opening or full closing of the valve, failing to meet continuous regulation requirements. While regulating actuators can receive standard analog commands for opening control, their core control algorithms largely rely on traditional proportional-integral-derivative (PID) control strategies. However, actual pipeline systems often exhibit complex characteristics such as strong nonlinearity, time-varying nature, and multivariable coupling. First, there is no one-to-one linear relationship between the opening degree of an electric butterfly valve and the flow rate and pressure of the medium. Because the pressure difference across the valve fluctuates dynamically with changes in opening degree and operating conditions, the flow gain is difficult to calculate accurately, making it difficult for linear control methods based on fixed models to maintain good performance across the entire operating range. Second, pipeline operation involves the mutual coupling and interference of multiple physical quantities. For example, parameters such as medium pressure, flow rate, temperature, and pressure difference across the valve are interrelated. When regulating the butterfly valve to change a parameter, it often triggers a chain reaction of fluctuations in other parameters. Traditional control methods often use independent adjustment of single parameters, failing to consider the coupling effects between variables, easily leading to overshoot, oscillation, or even instability. In addition, pipeline operating conditions change frequently in industrial production, such as sudden load changes, medium switching, and ambient temperature fluctuations. The optimal adjustment parameters for different operating conditions vary significantly, and the existing system lacks the ability to identify and adaptively match operating conditions, resulting in a decrease in control performance as the operating conditions deviate.

[0004] Besides the limitations of control strategies, changes in the state of the electric butterfly valve actuator itself are also a key factor affecting long-term accuracy. With increased operating time, mechanical wear, sensor drift, and changes in ambient temperature can all lead to zero-position offset and stroke errors in the actuator, causing a discrepancy between the commanded opening and the actual valve plate position. Currently, conventional systems lack self-diagnosis and adaptive calibration mechanisms for the actuator's state, often requiring regular manual maintenance, making it difficult to guarantee adjustment accuracy throughout the entire lifespan.

[0005] In summary, existing electric butterfly valve control technology has significant shortcomings in dealing with issues such as multivariable coupling, adaptation to changing operating conditions, and aging actuators. A precise regulation system capable of multivariable collaborative sensing, intelligent decision-making, and adaptive execution is needed to improve the automation level and control quality of pipeline systems. Summary of the Invention

[0006] To address the aforementioned problems in the existing technology, this invention provides a precise adjustment system and method for an electric butterfly valve based on multi-variable pipeline coordination. The objective of this invention can be achieved through the following technical solutions: The precision control system for electric butterfly valves based on pipeline multivariate collaboration includes: a multivariate sensing unit, a collaborative decision-making unit, a butterfly valve execution unit, and a closed-loop optimization unit; The multivariate sensing unit acquires multivariate parameters of pipeline operation and performs noise reduction preprocessing to generate multivariate original feature data. The original feature data is then subjected to coupling correlation analysis and normalization processing to generate multivariate collaborative feature data. The collaborative decision-making unit constructs a multi-scenario pipeline operating condition benchmark database, retrieves the matching benchmark parameters and feature data deviations in the database to perform collaborative adjustment operations, and generates collaborative control parameters for butterfly valve opening and execution rate. The butterfly valve actuator unit performs adaptive zero-position and stroke calibration operations on the electric butterfly valve actuator; it adjusts the butterfly valve plate opening and actuator operating speed based on collaborative control parameters, and simultaneously collects and dynamically corrects the execution parameters in real time. The closed-loop optimization unit generates standardized feedback feature data through multivariate coupling deviation separation, and performs quantitative verification of the feedback feature data and calibration benchmark parameters to form a closed-loop iterative feedback link.

[0007] Specifically, the process of generating multivariate original feature data is as follows: The acquired multivariate parameters of pipeline operation are synchronized and aligned with signals. An adaptive threshold noise reduction mechanism is adopted to identify and filter invalid signals generated by pipeline vibration, environmental electromagnetic interference and sensor drift, generate characteristic component data corresponding to each variable, and establish a linkage between noise reduction effect and load to ensure noise reduction accuracy under different working conditions. The amplitude of each variable parameter after filtering is calibrated to generate multivariate raw feature data.

[0008] Specifically, the process of performing coupling correlation analysis and normalization is as follows: The Pearson correlation coefficient combined with the mutual information method was used to perform coupling correlation analysis on the parameters in the multivariate original feature data, quantify the correlation strength between different parameters, and identify and screen the coupling feature pairs that affect the butterfly valve regulation. Each feature parameter is mapped to the same standardized interval, and the results of the coupling analysis are integrated with the normalized parameters to generate multivariate collaborative feature data.

[0009] Specifically, the process of constructing the multi-scenario pipeline operating condition benchmark database is as follows: Based on pipeline operating load, medium characteristics and environmental conditions, pipeline operating conditions are hierarchically clustered to form a set of multiple typical operating conditions. Collect multivariate parameters of pipelines throughout the entire cycle under various typical operating conditions, and extract the steady-state characteristic benchmark value and dynamic fluctuation threshold of each operating condition; By introducing a working condition similarity matching algorithm, a mapping model between working condition features and butterfly valve adjustment parameters is established. Feature benchmark values, fluctuation thresholds and corresponding adjustment parameters are associated and archived. A hierarchical index is constructed according to working condition categories to generate a multi-scenario pipeline working condition benchmark database.

[0010] Specifically, the process of introducing the working condition similarity matching algorithm is as follows: The K-nearest neighbor similarity algorithm is used to extract the feature benchmark values ​​of various working conditions in the multi-scenario pipeline working condition benchmark database as comparison samples; Calculate the similarity coefficient between the current feature data and each sample, and select the corresponding working condition category, feature benchmark value and adjustment parameter; At the same time, a similarity threshold is set to perform hierarchical comparison of the similarity between the current feature data and the database samples, and to perform accurate matching of working conditions and marking of unknown working conditions based on the comparison results.

[0011] Specifically, the process of performing the coordinated adjustment operation is as follows: The benchmark parameters matched in the multi-scenario pipeline operating condition benchmark database are retrieved and compared with the current multivariate collaborative feature data to classify and extract the deviation values ​​and deviation change trends of each parameter. A deviation weight allocation mechanism is introduced, which assigns different weights to each parameter based on its influence on the butterfly valve's adjustment accuracy, and calculates the deviation values ​​by weight to generate a comprehensive deviation value. By combining the comprehensive deviation value and the deviation change trend, the adjustment gradient of the butterfly valve opening degree and the execution rate is dynamically adjusted to generate the coordinated control parameters of the butterfly valve opening degree and the execution rate.

[0012] Specifically, the deviation weight allocation mechanism includes: The influence coefficients of various pipeline parameters on the butterfly valve adjustment accuracy were determined by experimentation, and initial weight values ​​were set for each parameter in combination with pipeline operation safety priority. Real-time acquisition of deviation correction effect data during the adjustment process, and dynamic iterative calibration of the initial weight value; The calibrated weight values ​​are added to the collaborative adjustment logic to establish a corresponding relationship between the weight values ​​and the comprehensive deviation value calculation. The weight values ​​are then called to perform a weighted calculation of the comprehensive deviation value, and the weight value call log and calculation results are recorded synchronously.

[0013] Specifically, the adaptive zero-position and travel calibration operation is performed as follows: The electric butterfly valve actuator is controlled to drive the valve plate to the mechanical limit position, and the feedback signal of the actuator at the mechanical limit position is collected to generate an initial zero position reference. The valve plate is driven to perform a full-stroke reciprocating motion, and feedback signals from multiple detection points within the stroke are collected and compared with the preset stroke threshold range. At the same time, the comparison threshold is dynamically adjusted according to the current working conditions. Calculate the stroke compensation amount based on the comparison and verification results, and generate a calibration parameter table adapted to the operating status of the butterfly valve.

[0014] Specifically, the process of regulating the opening degree of the butterfly valve plate and the operating speed of the actuator is as follows: Obtain the target opening value and the reference rate value from the cooperative control parameters, and decompose the target opening value into multi-level gradient sub-objectives; Based on the spacing of gradient sub-objectives, the corresponding execution rate sub-values ​​are matched to construct a gradient control sequence; The control actuator adjusts the opening degree of each sub-target sequentially according to the gradient control sequence, and dynamically corrects the execution rate sub-value by combining the real-time position feedback signal of the valve plate.

[0015] Specifically, the process of real-time acquisition and dynamic correction of execution parameters is as follows: The system acquires execution parameters such as the current valve opening and the real-time operating speed of the actuator. The execution parameters are compared and verified with the target parameters in the collaborative control parameters, and the parameter deviation is extracted. The correction level is determined based on the magnitude of the deviation, and a corresponding real-time correction amount is generated to form a real-time closed-loop control mechanism for the execution process.

[0016] Specifically, the process of generating standardized feedback feature data is as follows: The pipeline multivariable parameters after butterfly valve adjustment are obtained, and the pipeline multivariable parameters are subjected to signal synchronous calibration. Through parameter filtering and amplitude normalization operations, parameter fluctuation correction values ​​are generated. After calibration, deviation extraction, classification, and grading screening operations are performed on the parameters to divide the adjustment deviation components of each parameter. The separated deviation components are normalized and calibrated to generate standardized feedback characteristic data.

[0017] Specifically, the quantitative verification process is as follows: Standardized feedback feature data and calibration benchmark parameters are retrieved, and a dynamic allocation mechanism for verification weights is adopted. The allocation ratio of verification weights for each parameter is dynamically adapted in combination with the real-time operating load of the pipeline. Based on the weight allocation results, differentiated comparison and calculation are performed on each parameter, and a comprehensive verification coefficient and deviation contribution ratio are generated simultaneously. The linkage between the verification results and the operating load is established, and a graded verification judgment standard is generated. Based on dynamic judgment criteria, the system completes the calibration of each parameter to meet the standards, locates the causes of deviations and quantifies their impact, and generates a quantitative verification log and correction instructions with priority sorting.

[0018] The system achieves end-to-end adaptive control through the deep linkage of four major units, namely "multi-variable acquisition - collaborative analysis - working condition matching - precise execution - closed-loop optimization". It breaks through the limitations of single-link optimization and independent operation of each unit in the existing technology, and solves the technical problems of low adjustment accuracy and poor adaptability of existing electric butterfly valves.

[0019] Specifically, the precise adjustment method for electric butterfly valves based on multi-variable pipeline coordination includes: S1: Obtain multivariate parameters of pipeline operation, perform signal synchronization alignment, adaptive threshold noise reduction and amplitude calibration on the parameters to generate multivariate raw feature data; perform coupling correlation analysis and normalization processing on the multivariate raw feature data to generate multivariate collaborative feature data; S2: Construct a multi-scenario pipeline operating condition benchmark database, perform coordinated adjustment operations based on the benchmark parameters matched in the database and multi-variable coordinated feature data, introduce a deviation weight allocation mechanism to perform weighted accounting of the deviations of each parameter, and generate coordinated control parameters for butterfly valve opening and execution rate; S3: Perform adaptive zero-position and stroke calibration on the electric butterfly valve actuator, adjust the butterfly valve plate opening and actuator operating speed according to the coordinated control parameters, and collect and dynamically correct the execution parameters in real time. S4: Based on multivariate coupling deviation separation processing, standardized feedback feature data is generated. The standardized feedback feature data and calibration benchmark parameters are quantitatively verified to form a closed-loop iterative feedback link.

[0020] The beneficial effects of this invention are as follows: (1) By using a multivariable sensing unit to achieve coordinated acquisition and noise reduction of signals, the accuracy and reliability of the original data are improved. This invention sets up a multivariable sensing unit to synchronously acquire and align signals of multiple sources such as pressure, flow rate, temperature, and differential pressure during pipeline operation, thus solving the control deviation problem caused by asynchronous time of multi-source data. An adaptive threshold noise reduction mechanism is adopted to intelligently identify and filter invalid signals generated by pipeline vibration, environmental electromagnetic interference, and sensor drift, and dynamically adjust the noise reduction parameters according to the operating load to ensure noise reduction accuracy under different operating conditions. The filtered parameters are amplitude calibrated to generate high-quality multivariable original feature data, providing a reliable data foundation for subsequent precise adjustment and effectively avoiding erroneous adjustment caused by data distortion.

[0021] (2) Intelligent collaborative decision-making under multivariate coupled interference is achieved through coupling correlation analysis and operating condition matching mechanism. This invention uses Pearson correlation coefficient combined with mutual information method to perform coupling correlation analysis on multivariate original feature data, quantifies the correlation strength between different parameters, and identifies and screens out key coupling feature pairs affecting butterfly valve regulation. After normalizing each feature parameter, the coupling analysis results are integrated to generate multivariate collaborative feature data. The collaborative decision-making unit constructs a multi-scenario pipeline operating condition benchmark database, identifies the current operating condition in real time through operating condition similarity matching algorithm, retrieves the matched benchmark parameters and compares the deviation with the current feature data, introduces a deviation weight dynamic allocation mechanism to perform weighted accounting of the deviation of each parameter, and generates collaborative control parameters for butterfly valve opening and execution rate, realizing precise collaborative control under multivariate coupled operating conditions.

[0022] (3) High-reliability and precise adjustment throughout the entire lifecycle is achieved through adaptive calibration and closed-loop iterative optimization of the actuator. This invention sets up a butterfly valve actuator unit, performing adaptive zero-position and stroke calibration on the electric butterfly valve actuator before adjustment. Through mechanical limit detection and comparison with multi-point feedback throughout the stroke, the stroke compensation is calculated to generate a calibration parameter table, effectively eliminating zero-position deviation caused by mechanical wear and sensor drift during long-term operation. During the control process, the target opening value is decomposed into multi-level gradient sub-targets and matched with corresponding execution rates. Combined with real-time position feedback for dynamic correction, adjustment shock is avoided. The closed-loop optimization unit performs deviation separation and quantification verification on the adjusted parameters, forming a closed-loop iterative feedback link. Through continuous self-optimization, the equipment maintenance cycle is extended, and manual maintenance costs are reduced. Attached Figure Description

[0023] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0024] Figure 1 This is a schematic diagram of a precise adjustment system and method for an electric butterfly valve based on multi-variable pipeline coordination according to the present invention. Figure 2 This is a timing diagram of a precision adjustment system for an electric butterfly valve based on multi-variable pipeline coordination, as described in this invention. Detailed Implementation

[0025] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0026] Please see Figure 1-2 A precision control system for an electric butterfly valve based on pipeline multivariate coordination includes: a multivariate sensing unit, a collaborative decision-making unit, a butterfly valve execution unit, and a closed-loop optimization unit; The multivariate sensing unit acquires multivariate parameters of pipeline operation and performs noise reduction preprocessing to generate multivariate original feature data. The original feature data is then subjected to coupling correlation analysis and normalization processing to generate multivariate collaborative feature data. The multivariate parameters of pipeline operation include pipeline medium pressure, flow rate, medium viscosity, and pipeline temperature. The collaborative decision-making unit constructs a multi-scenario pipeline operating condition benchmark database, retrieves the matching benchmark parameters and feature data deviations in the database to perform collaborative adjustment operations, and generates collaborative control parameters for butterfly valve opening and execution rate. The butterfly valve actuator unit performs adaptive zero-position and stroke calibration operations on the electric butterfly valve actuator; it adjusts the butterfly valve plate opening and actuator operating speed based on collaborative control parameters, and simultaneously collects and dynamically corrects the execution parameters in real time. The closed-loop optimization unit generates standardized feedback feature data through multivariate coupling deviation separation, and performs quantitative verification of the feedback feature data and calibration benchmark parameters to form a closed-loop iterative feedback link.

[0027] Specifically, the process of generating multivariate original feature data is as follows: An adaptive threshold noise reduction mechanism is adopted to identify and filter invalid signals generated by pipeline vibration, environmental electromagnetic interference and sensor drift, generate characteristic component data corresponding to each variable, and establish a linkage between noise reduction effect and load to ensure noise reduction accuracy under different working conditions. The amplitude of each variable parameter after filtering is calibrated to generate multivariate original characteristic data.

[0028] In this embodiment, the pipeline system is applied to an industrial circulating water transportation scenario, and during operation, there are interference factors such as pipeline vibration, electromagnetic interference, and sensor temperature drift. After the multivariable sensing unit collects multivariable parameters such as pressure, flow rate, and temperature in real time, it first performs signal synchronization and alignment of each parameter to ensure that the parameters of different acquisition channels are consistent in timing.

[0029] Subsequently, an adaptive threshold noise reduction mechanism is activated, dynamically adjusting the noise reduction threshold range based on the real-time operating load of the pipeline. Under low-load stable operating conditions, the threshold range adaptively narrows to finely filter minute fluctuation signals; under high-load fluctuating conditions, the threshold range adaptively widens to retain effective fluctuation characteristics and eliminate invalid signals generated by vibration, electromagnetic interference, and sensor drift. By comparing the acquired signals with the dynamic threshold in real time, invalid signals are identified and filtered, generating characteristic component data corresponding to each variable, and establishing a linkage between the noise reduction effect and the operating load, enabling the noise reduction processing to adapt to different operating conditions.

[0030] The filtered feature component data are calibrated for amplitude. The signal amplitude is normalized and corrected according to the sensor's rated output range to eliminate amplitude offset caused by the acquisition link. Finally, stable and uniform multivariate original feature data are generated, providing a reliable data foundation for subsequent coupled correlation analysis.

[0031] Specifically, the process of performing coupling correlation analysis and normalization is as follows: The Pearson correlation coefficient combined with mutual information method is used to perform coupling correlation analysis on the parameters in the original multivariate feature data, quantifying the correlation strength between different parameters and identifying and screening coupling feature pairs that affect butterfly valve regulation. A normalization algorithm is used to map each feature parameter to the same standardized interval, eliminating the dimensional differences between different parameters. The coupling analysis results and the normalized parameters are integrated to generate multivariate collaborative feature data that can be directly used for collaborative decision-making, ensuring the consistency and correlation of the feature data.

[0032] In this embodiment, based on an industrial circulating water pipeline scenario, the multivariate sensing unit has generated multivariate raw feature data, including pipeline pressure. Flow rate Medium temperature Differential pressure before and after the valve The four core feature parameters, each of which are valid data after adaptive threshold denoising and amplitude calibration, are then subjected to coupling correlation analysis and normalization processing, as detailed below: First, a coupling correlation analysis was performed on the four parameters in the original multivariate feature data using the Pearson correlation coefficient combined with mutual information to quantify the correlation strength between different parameters. The Pearson correlation coefficient is used to quantify the degree of linear correlation between the parameters, and its calculation formula is as follows: In the formula, Let be the Pearson correlation coefficient between the i-th parameter and the j-th parameter, with a value range of [-1, 1]. This refers to the k-th collected data for the i-th parameter; is the average value of all collected data for the i-th parameter; n is the total amount of collected data for a single parameter.

[0033] Mutual information is used to supplement the quantification of the nonlinear correlation between parameters, overcoming the limitation that the Pearson correlation coefficient can only reflect linear correlations. Its calculation formula is as follows: , In the formula, ; Let be the mutual information value between the i-th parameter and the j-th parameter. The larger the mutual information value, the stronger the nonlinear correlation between the two parameters. , The marginal probability densities of the i-th and j-th parameters are respectively; p(xi, Let be the joint probability density of the i-th parameter and the j-th parameter.

[0034] Combining Pearson correlation coefficient With mutual information value ; Construct a comprehensive correlation strength evaluation index The calculation formula is: ,in These are weighting coefficients (ranging from [0,1]), used to balance the effects of linear and nonlinear associations. The maximum value among all parameter pairs' mutual information values ​​is used to normalize the mutual information values, making them... The value range is uniformly [0,1].

[0035] Set a comprehensive correlation strength threshold ,when When the parameter pair is determined to be a coupling feature pair affecting the butterfly valve regulation, it is identified and screened; when When the parameter is determined to have a weak correlation and does not affect the butterfly valve adjustment, it is discarded. In this embodiment, the parameters obtained after screening are... , (Pressure and pressure difference) , The two sets of core coupling features (flow rate and pressure difference) provide the core correlation basis for subsequent coordinated regulation.

[0036] Subsequently, a minimum-maximum normalization algorithm is used to map each feature parameter to the same standardized interval [a, b], eliminating the dimensional differences between different parameters (e.g., pressure is in MPa, flow rate is in m³). 3 (The unit of temperature is °C, and the dimensions are not uniform), the calculation formula is: In the formula, The normalized value of the i-th parameter; The minimum value in the original feature data of the i-th parameter; is the maximum value in the original feature data of the i-th parameter; a and b are the upper and lower limits of the standardized interval (which can be set according to the actual pipeline operating conditions).

[0037] Finally, the core coupling feature pairs selected by the above coupling correlation analysis, along with the normalized feature parameters, are integrated. , , , The correlation strength information of coupled feature pairs is linked and bound with normalized parameters to generate multivariate collaborative feature data that can be directly used by collaborative decision-making units for working condition matching and collaborative adjustment operations. This ensures the consistency and correlation of feature data and provides accurate data support for the subsequent generation of collaborative control parameters.

[0038] Specifically, the process of constructing the multi-scenario pipeline operating condition benchmark database is as follows: First, based on pipeline operating load, media characteristics, and environmental conditions, pipeline operating conditions are hierarchically clustered to distinguish the core feature differences of different operating conditions, forming multiple typical operating condition sets. Then, multivariate pipeline parameters throughout the entire cycle under each typical operating condition are collected. After data purification using the preceding processing steps, steady-state characteristic benchmark values ​​and dynamic fluctuation thresholds for each operating condition are extracted. An operating condition similarity matching algorithm is introduced to establish a mapping model between operating condition features and butterfly valve adjustment parameters. The characteristic benchmark values, fluctuation thresholds, and corresponding adjustment parameters are associated and archived. A hierarchical index is constructed according to operating condition categories, ultimately generating a multi-scenario pipeline operating condition benchmark database that enables rapid matching and retrieval, breaking through the limitations of traditional databases that simply archive data.

[0039] Specifically, the introduced working condition similarity matching algorithm is implemented as follows: The K-nearest neighbor similarity algorithm is used, with the current pipeline multivariate collaborative feature data as the retrieval core. Feature benchmark values ​​for various working conditions in the database are extracted as comparison samples. The similarity coefficient between the current feature data and each sample is calculated, and the working condition category with the highest similarity and its corresponding feature benchmark value and adjustment parameters are selected. Simultaneously, a similarity threshold is set, and the similarity between the current feature data and the database samples is compared hierarchically. Based on the comparison results, precise working condition matching or unknown working condition labeling is achieved. During precise matching, the corresponding adjustment parameters are retrieved for collaborative decision-making. When an unknown working condition is labeled, the data is temporarily stored for database iteration, ensuring the accuracy of working condition matching and the continuity of database updates. This provides incremental data for subsequent database iteration updates, achieving both accuracy and flexibility in working condition matching. When the highest similarity does not reach the threshold, it is automatically labeled as an unknown working condition, and the corresponding data is temporarily stored.

[0040] Specifically, the process of performing the coordinated adjustment operation is as follows: The benchmark parameters matched in the multi-scenario pipeline operating condition benchmark database are retrieved and compared with the current multivariate collaborative feature data to classify and extract the deviation values ​​and deviation change trends of each parameter. A deviation weight allocation mechanism is introduced, which assigns different weights to each parameter based on its influence on the butterfly valve's adjustment accuracy, and calculates the deviation values ​​by weight to generate a comprehensive deviation value. By combining the comprehensive deviation value and the deviation change trend, the adjustment gradient of the butterfly valve opening and execution rate is dynamically adjusted to generate collaborative control parameters adapted to the current working conditions, thereby achieving precise adjustment under multi-variable collaboration, which is different from the traditional single deviation adjustment mode.

[0041] Specifically, the deviation weight allocation mechanism includes: The influence coefficients of various pipeline parameters on the butterfly valve adjustment accuracy were determined by experimentation, and initial weight values ​​were set for each parameter in combination with pipeline operation safety priority. Real-time acquisition of deviation correction effect data during the adjustment process, and dynamic iterative calibration of the initial weight value; The calibrated weight values ​​are added to the collaborative adjustment logic to establish a corresponding relationship between the weight values ​​and the comprehensive deviation value calculation. The weight values ​​are called to perform the weighted calculation of the comprehensive deviation value, and the weight value call log and calculation results are recorded synchronously for subsequent secondary calibration of the weight values.

[0042] Specifically, the adaptive zero-position and travel calibration operation is performed as follows: The electric butterfly valve actuator is controlled to drive the valve plate to the mechanical limit position, and the feedback signal of the actuator at the mechanical limit position is collected to generate an initial zero position reference. The valve plate is driven to perform a full-stroke reciprocating motion, and feedback signals from multiple detection points within the stroke are collected and compared with the preset stroke threshold range. At the same time, the comparison threshold is dynamically adjusted according to the current working conditions to achieve linkage adaptation between calibration and pipeline working conditions, thereby improving calibration accuracy under different working conditions. The stroke compensation amount is calculated based on the comparison and verification results. The initial zero position reference and stroke compensation amount are entered into the actuator control module to generate a calibration parameter table that adapts to the current butterfly valve operating status, thus completing the full-process calibration operation of adaptive zero position and stroke.

[0043] Specifically, the process of regulating the opening degree of the butterfly valve plate and the operating speed of the actuator is as follows: Receive the opening target value and the rate reference value from the cooperative control parameters, and decompose the opening target value into multi-level gradient sub-targets; Based on the spacing of gradient sub-objectives, the corresponding execution rate sub-values ​​are matched to construct a gradient control sequence; The control actuator adjusts the opening degree of each sub-target sequentially according to the gradient control sequence, and dynamically corrects the execution rate sub-value by combining the real-time position feedback signal of the valve plate, so as to realize the synchronous and coordinated control of the valve plate opening degree and the operating speed of the actuator.

[0044] In this embodiment, based on the scenario of an industrial circulating water pipeline, the multivariate collaborative feature data (including the normalized pipeline pressure) generated by the preceding coupling correlation analysis and normalization processing is used. Flow rate Medium temperature Differential pressure before and after the valve and the selected coupling feature pairs , The collaborative decision-making unit executes collaborative adjustment operations, and at the same time introduces a deviation weight allocation mechanism to complete weight setting, calibration and application. The specific process is as follows, ensuring that each step of the operation can be implemented and is fully transparent.

[0045] I. Perform coordinated adjustment operations The first step is to retrieve benchmark parameters and compare deviations. The collaborative decision-making unit first retrieves the existing multi-scenario pipeline operating condition benchmark database. Based on the operating condition similarity matching algorithm mentioned above, it filters out parameters that match the current pipeline operating condition (circulating water operating at low load and stable, medium temperature stable at...). The nearest matching set of reference parameters, whose corresponding normalized reference values ​​are as follows: (Pressure reference value) (Flow rate baseline value) (Temperature reference value) (Pressure difference baseline value). Subsequently, the deviation values ​​of each parameter in the current multivariate co-variable feature data are compared one-to-one with the corresponding baseline values, and the deviation values ​​of each parameter are extracted by classification using the difference method. The calculation formula is: Simultaneously, five consecutive sets of current collaborative feature data are collected, and the deviation trend of each parameter is extracted by calculating the difference in deviation values ​​between adjacent sets of data. ,when When the judgment deviation shows an increasing trend, When the judgment deviation shows a decreasing trend, At that time, the judgment deviation remained stable.

[0046] The second step involves introducing a deviation weighting allocation mechanism to calculate the overall deviation value using weighted calculations. Based on the degree of influence of each parameter on the butterfly valve's adjustment accuracy, different weights are assigned to each parameter's deviation value through the deviation weighting allocation mechanism. A weighted calculation is performed to generate a comprehensive deviation value ∆Y. The specific weight allocation and calculation process are detailed in Part II of this embodiment. The formula for calculating the comprehensive deviation value is as follows: , in Let each parameter be a weight value, and satisfy the following conditions: .

[0047] The third step involves dynamically adjusting the control gradient to generate coordinated control parameters. This is combined with the calculated comprehensive deviation value ∆Y and the deviation trends of each parameter. Dynamically adjust the adjustment gradient between the butterfly valve opening degree and the actuator operating speed: when ( (for setting a high deviation threshold) and At that time, the butterfly valve opening adjustment gradient is increased to Execution rate adjustment gradient increased to To accelerate the adjustment speed and quickly reduce the deviation; when ( When the preset low deviation threshold is met, adjust the gradient to the intermediate level. , It balances adjustment speed and precision; when and When the gradient is adjusted to decrease , To avoid over-adjustment that could lead to a reverse increase in deviation, a target butterfly valve opening value suitable for the current low-load stable operating condition is generated based on the adjusted adjustment gradient. The target value v of the actuator's operating speed is sent to the butterfly valve actuator as a coordinated control parameter to achieve precise adjustment under multi-variable coordination, which is different from the traditional single deviation adjustment mode (adjustment is based only on the deviation of a single parameter, ignoring the coupling relationship between parameters and the trend of deviation change).

[0048] II. Specific Implementation of the Deviation Weight Allocation Mechanism The first step is to calibrate the influence coefficients and set initial weight values. The influence coefficients of various pipeline parameters on the butterfly valve's adjustment accuracy are calibrated using orthogonal experimental design. During the experiment, the pressure is changed individually. ,flow ,temperature Pressure difference The parameter values ​​were recorded, the changes in the butterfly valve's adjustment accuracy were recorded, and the influence coefficients of each parameter were calculated. (Pressure Influence Coefficient) (Flow Influence Coefficient) (Temperature Influence Coefficient) (Differential pressure influence coefficient) The larger the influence coefficient, the more significant the impact of this parameter on the butterfly valve's adjustment accuracy. Based on the pipeline operation safety priority (differential pressure and pressure directly affect pipeline safety, so they have the highest priority; flow rate is next; temperature has the lowest priority), initial weight values ​​are set for each parameter. The initial weight values ​​are set by combining the influence coefficient and safety priority, and are obtained through normalization to ensure that the sum of the initial weight values ​​is 1. In this embodiment, the initial weight values ​​are set as follows: , , , ,in This is consistent with the high priority and high influence coefficient of pressure difference and pressure.

[0049] The second step is to dynamically iteratively calibrate the initial weight values. After the butterfly valve actuator performs the adjustment operation based on the coordinated control parameters, it collects the deviation correction effect data in real time during the adjustment process, that is, the change in the deviation value of each parameter before and after adjustment. (The difference between the adjusted deviation value and the original deviation value), the evaluation index of deviation correction effect is calculated. ( The larger the value, the better the deviation correction effect of the parameter. Dynamic iterative calibration is performed on the initial weight values. If a parameter's... Less than the preset correction effect threshold This indicates that the weighting of this parameter is unreasonable and its weight value needs to be increased appropriately; if If the weight value remains unchanged or is slightly adjusted, then the calibration formula is as follows: ,in The calibration coefficient (ranging from [0,1]) is used to control the magnitude of weight calibration and avoid sudden weight changes affecting the stability of the adjustment. After calibration, all weight values ​​are normalized to ensure that their sum is still 1, thus obtaining the calibrated weight values. .

[0050] The third step is the application and logging of the weight values. The calibrated weight values ​​will be... Add it to the collaborative adjustment logic to establish the correspondence between weight values ​​and comprehensive deviation value calculation, and clarify each weight value. Deviation value from corresponding parameters The matching relationship is used to automatically call the corresponding weight value each time the comprehensive deviation value is calculated. Substitute into the formula: The weighted calculation is then completed. Simultaneously, a weight value call log is recorded, including the call time and the weight value called. , corresponding parameter deviation values Comprehensive deviation value ∆Y and deviation correction effect The log data is used for subsequent secondary calibration of the weight values. Each secondary calibration is based on the most recent set of log data and dynamically adjusts the weight values ​​in combination with the deviation correction effect to ensure that the deviation weight allocation mechanism always adapts to the current pipeline operating conditions and improves the accuracy of coordinated adjustment.

[0051] Specifically, the process of real-time acquisition and dynamic correction of execution parameters is as follows: The system acquires execution parameters such as the current valve opening and the real-time operating speed of the actuator. The execution parameters are compared and verified with the target parameters in the collaborative control parameters, and the parameter deviation is extracted. The correction level is determined based on the deviation amplitude (e.g., deviation ≤ 5% is Level 1 correction, 5% < deviation ≤ 10% is Level 2 correction), and a corresponding real-time correction amount is generated according to the correction level. A correction is completed once every preset time interval, forming a real-time closed-loop control mechanism for the execution process.

[0052] Specifically, the process of generating standardized feedback feature data is as follows: After the butterfly valve is adjusted, the multivariable parameters of the pipeline are acquired, and the multivariable parameters of the pipeline are subjected to signal synchronous calibration. Through parameter filtering and amplitude normalization, parameter fluctuation correction values ​​are generated for calibration of the original acquired parameters. A multivariable coupling deviation separation algorithm is used to extract, classify and grade the deviation of the calibrated parameters to divide the adjustment deviation components of each parameter. The separated deviation components are normalized and calibrated, and the normalized deviation components are integrated to generate standardized feedback feature data, which serves as the basis for subsequent quantitative verification.

[0053] Specifically, the quantitative verification process is as follows: Standardized feedback feature data and calibration benchmark parameters are retrieved, and a dynamic weight allocation mechanism for verification is adopted. Combined with the real-time operating load of the pipeline, the weight allocation ratio of each parameter is dynamically adapted (the correlation logic between weight adaptation and operating load is: the higher the operating load, the greater the impact of the butterfly valve adjustment accuracy on the parameter, the higher the weight allocation ratio, and vice versa). Based on the weight allocation results, differentiated comparison and calculation are performed on each parameter, and a comprehensive verification coefficient and deviation contribution ratio are generated simultaneously. The linkage between the verification results and the operating load is established, and a graded verification judgment standard is generated. Based on dynamic judgment criteria, the system completes the calibration of each parameter to meet the standards, locates the causes of deviations and quantifies their impact, generates a quantitative verification log with priority sorting and correction instructions, and realizes dynamic self-adaptation of the verification process.

[0054] The system achieves end-to-end adaptive control through the deep linkage of four major units, namely "multi-variable acquisition - collaborative analysis - working condition matching - precise execution - closed-loop optimization". It breaks through the limitations of single-link optimization and independent operation of each unit in the existing technology, and solves the technical problems of low adjustment accuracy and poor adaptability of existing electric butterfly valves.

[0055] Specifically, a method for precise adjustment of an electric butterfly valve based on multi-variable pipeline coordination includes: S1: Obtain multivariate parameters of pipeline operation, perform signal synchronization alignment, adaptive threshold noise reduction and amplitude calibration on the parameters to generate multivariate raw feature data; perform coupling correlation analysis and normalization processing on the multivariate raw feature data to generate multivariate collaborative feature data; S2: Construct a multi-scenario pipeline operating condition benchmark database, perform coordinated adjustment operations based on the benchmark parameters matched in the database and multi-variable coordinated feature data, introduce a deviation weight allocation mechanism to perform weighted accounting of the deviations of each parameter, and generate coordinated control parameters for butterfly valve opening and execution rate; S3: Perform adaptive zero-position and stroke calibration on the electric butterfly valve actuator, adjust the butterfly valve plate opening and actuator operating speed according to the coordinated control parameters, and collect and dynamically correct the execution parameters in real time. S4: Based on multivariate coupling deviation separation processing, standardized feedback feature data is generated. The standardized feedback feature data and calibration benchmark parameters are quantitatively verified to form a closed-loop iterative feedback link.

[0056] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A precision control system for electric butterfly valves based on multi-variable pipeline coordination, characterized in that, include: Multivariable sensing unit, collaborative decision-making unit, butterfly valve execution unit, closed-loop optimization unit; The multivariate sensing unit acquires multivariate parameters of pipeline operation and performs noise reduction preprocessing to generate multivariate original feature data. The original feature data is then subjected to coupling correlation analysis and normalization processing to generate multivariate collaborative feature data. The collaborative decision-making unit constructs a multi-scenario pipeline operating condition benchmark database, retrieves the matching benchmark parameters and feature data deviations in the database to perform collaborative adjustment operations, and generates collaborative control parameters for butterfly valve opening and execution rate. The butterfly valve actuator unit performs adaptive zero-position and stroke calibration operations on the electric butterfly valve actuator. The butterfly valve plate opening degree and actuator operating speed are adjusted based on the coordinated control parameters, and the execution parameters are collected and dynamically corrected in real time. The closed-loop optimization unit generates standardized feedback feature data by separating multivariable coupling deviations, and performs quantitative verification of the feedback feature data and calibration benchmark parameters to form a closed-loop iterative feedback link. The specific process for constructing the multi-scenario pipeline operating condition benchmark database is as follows: Based on pipeline operating load, medium characteristics and environmental conditions, pipeline operating conditions are hierarchically clustered to form a set of multiple typical operating conditions. Collect multivariate parameters of pipelines throughout the entire cycle under various typical operating conditions, and extract the steady-state characteristic benchmark value and dynamic fluctuation threshold of each operating condition; A similarity matching algorithm for operating conditions is introduced to establish a mapping model between operating condition features and butterfly valve adjustment parameters. Feature benchmark values, fluctuation thresholds and corresponding adjustment parameters are associated and archived. A hierarchical index is constructed according to the operating condition category to generate a multi-scenario pipeline operating condition benchmark database. The specific process of introducing the working condition similarity matching algorithm is as follows: The K-nearest neighbor similarity algorithm is used to extract the feature benchmark values ​​of various working conditions in the multi-scenario pipeline working condition benchmark database as comparison samples; Calculate the similarity coefficient between the current feature data and each sample, and select the corresponding working condition category, feature benchmark value and adjustment parameter; At the same time, a similarity threshold is set to perform hierarchical comparison of the similarity between the current feature data and the database samples, and to perform accurate matching of working conditions and marking of unknown working conditions based on the comparison results; The specific process of performing the coordinated adjustment operation is as follows: The benchmark parameters matched in the multi-scenario pipeline operating condition benchmark database are retrieved and compared with the current multivariate collaborative feature data to classify and extract the deviation values ​​and deviation change trends of each parameter. A deviation weight allocation mechanism is introduced, which assigns different weights to each parameter based on its influence on the butterfly valve's adjustment accuracy, and calculates the deviation values ​​by weight to generate a comprehensive deviation value. By combining the comprehensive deviation value and the deviation change trend, the adjustment gradient of the butterfly valve opening degree and the execution rate is dynamically adjusted to generate the coordinated control parameters of the butterfly valve opening degree and the execution rate. The deviation weight allocation mechanism includes: The influence coefficients of various pipeline parameters on the butterfly valve adjustment accuracy were determined by experimentation, and initial weight values ​​were set for each parameter in combination with pipeline operation safety priority. Real-time acquisition of deviation correction effect data during the adjustment process, and dynamic iterative calibration of the initial weight value; The calibrated weight values ​​are added to the collaborative adjustment logic to establish a corresponding relationship between the weight values ​​and the comprehensive deviation value calculation. The weight values ​​are then called to perform a weighted calculation of the comprehensive deviation value, and the weight value call log and calculation results are recorded synchronously.

2. The system according to claim 1, characterized in that, The specific process for generating multivariate original feature data is as follows: The acquired multivariate parameters of pipeline operation are synchronized and aligned with signals. An adaptive threshold noise reduction mechanism is adopted to identify and filter invalid signals generated by pipeline vibration, environmental electromagnetic interference and sensor drift, generate characteristic component data corresponding to each variable, and establish a linkage between noise reduction effect and load. The amplitude of each variable parameter after filtering is calibrated to generate multivariate raw feature data.

3. The system according to claim 1, characterized in that, The specific process of performing coupling correlation analysis and normalization is as follows: The Pearson correlation coefficient combined with the mutual information method was used to perform coupling correlation analysis on the parameters in the multivariate original feature data, quantify the correlation strength between different parameters, and identify and screen the coupling feature pairs that affect the butterfly valve regulation. Each feature parameter is mapped to the same standardized interval, and the results of the coupling analysis are integrated with the normalized parameters to generate multivariate collaborative feature data.

4. The system according to claim 1, characterized in that, The adaptive zero-position and travel calibration operation is performed as follows: The electric butterfly valve actuator is controlled to drive the valve plate to the mechanical limit position, and the feedback signal of the actuator at the mechanical limit position is collected to generate an initial zero position reference. The valve plate is driven to perform a full-stroke reciprocating motion, and feedback signals from multiple detection points within the stroke are collected and compared with the preset stroke threshold range. At the same time, the comparison threshold is dynamically adjusted according to the current working conditions. Calculate the stroke compensation amount based on the comparison and verification results, and generate a calibration parameter table adapted to the operating status of the butterfly valve.

5. The system according to claim 1, characterized in that, The specific process for regulating the opening degree of the butterfly valve plate and the operating speed of the actuator is as follows: Obtain the target opening value and the reference rate value from the cooperative control parameters, and decompose the target opening value into multi-level gradient sub-objectives; Based on the spacing of gradient sub-objectives, the corresponding execution rate sub-values ​​are matched to construct a gradient control sequence; The control actuator adjusts the opening degree of each sub-target sequentially according to the gradient control sequence, and dynamically corrects the execution rate sub-value by combining the real-time position feedback signal of the valve plate.

6. The system according to claim 1, characterized in that, The specific process of real-time acquisition and dynamic correction of execution parameters is as follows: The system acquires execution parameters such as the current valve opening and the real-time operating speed of the actuator. The execution parameters are compared and verified with the target parameters in the collaborative control parameters, and the parameter deviation is extracted. The correction level is determined based on the magnitude of the deviation, and a corresponding real-time correction amount is generated to form a real-time closed-loop control mechanism for the execution process.

7. The system according to claim 1, characterized in that, The specific process for generating standardized feedback feature data is as follows: The pipeline multivariable parameters after butterfly valve adjustment are obtained, and the pipeline multivariable parameters are subjected to signal synchronous calibration. Through parameter filtering and amplitude normalization operations, parameter fluctuation correction values ​​are generated. After calibration, deviation extraction, classification, and grading screening operations are performed on the parameters to divide the adjustment deviation components of each parameter. The separated deviation components are normalized and calibrated to generate standardized feedback characteristic data.

8. The system according to claim 1, characterized in that, The specific process for performing quantization verification is as follows: Standardized feedback feature data and calibration benchmark parameters are retrieved, and a dynamic allocation mechanism for verification weights is adopted. The allocation ratio of verification weights for each parameter is dynamically adapted in combination with the real-time operating load of the pipeline. Based on the weight allocation results, differentiated comparison and calculation are performed on each parameter, and a comprehensive verification coefficient and deviation contribution ratio are generated simultaneously. The linkage between the verification results and the operating load is established, and a graded verification judgment standard is generated. Based on dynamic judgment criteria, the system completes the calibration of each parameter to meet the standards, locates the causes of deviations and quantifies their impact, and generates a quantitative verification log and correction instructions with priority sorting.

9. A method for precise adjustment of an electric butterfly valve based on multi-variable pipeline coordination, applied to the system described in any one of claims 1 to 8, characterized in that, include: S1: Obtain multivariate parameters of pipeline operation, perform signal synchronization alignment, adaptive threshold noise reduction and amplitude calibration on the parameters to generate multivariate raw feature data; perform coupling correlation analysis and normalization processing on the multivariate raw feature data to generate multivariate collaborative feature data; S2: Construct a multi-scenario pipeline operating condition benchmark database, perform coordinated adjustment operations based on the benchmark parameters matched by the database and multi-variable coordinated feature data, introduce a deviation weight allocation mechanism to perform weighted calculation of the deviation of each parameter, and generate coordinated control parameters for butterfly valve opening degree and execution rate. S3: Perform adaptive zero-position and stroke calibration on the electric butterfly valve actuator, adjust the butterfly valve plate opening and actuator operating speed according to the coordinated control parameters, and collect and dynamically correct the execution parameters in real time. S4: Based on multivariate coupling deviation separation processing, standardized feedback feature data is generated. The standardized feedback feature data and calibration benchmark parameters are quantitatively verified to form a closed-loop iterative feedback link.

Images