High-precision electric calibration method and system for wedge flowmeter

Abstract

The application discloses a high-precision electric calibration method and system for a wedge flowmeter, comprising the following steps: step one, data acquisition and pretreatment; step two, dynamic working condition identification; step three, adaptive model updating; step four, multi-source data fusion calibration; and step five, closed-loop verification and compensation. The application realizes comprehensive and accurate data acquisition by collecting multiple types of signals in real time through a multi-dimensional sensor array and constructing a data set, thereby providing a good foundation for calibration; dynamic working condition identification is performed by using a support vector machine to trigger model updating, thereby giving the calibration dynamic adaptive capability; recursive least squares method is used to update model parameters and establish a compensation term, thereby enhancing the model representation capability; a data fusion algorithm is used to generate a calibration coefficient, thereby improving the calibration accuracy; and the calibration is verified and compensated through a digital twin model, thereby effectively overcoming the traditional defects, reducing model transplantation errors, improving the measurement accuracy of the wedge flowmeter under extreme working conditions, and guaranteeing production stability and product quality.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of flow calibration, in particular to a high-precision electric calibration method and system for wedge flow meters. BACKGROUND

[0002] As a key flow measurement instrument in industrial process control, the measurement accuracy of wedge flow meters directly affects the stability of the production process and product quality. With the continuous improvement of the measurement accuracy requirements of modern industry, traditional calibration techniques have been difficult to meet the high-precision measurement requirements under complex working conditions.

[0003] Currently, the calibration of wedge flow meters mainly relies on static calibration in a laboratory environment, and a fixed mathematical model is established to describe the relationship between flow and pressure difference. This method can achieve basic measurement functions under ideal working conditions, but in actual industrial scenarios, the dynamic changes in fluid temperature, pressure, viscosity and other parameters will significantly affect the measurement accuracy.

[0004] There are three defects in the prior art: first, the traditional calibration model lacks dynamic adaptive ability and cannot respond to changes in working condition parameters in real time; second, the single sensor data source limits the model's ability to characterize complex flow fields; third, the calibration process and the actual measurement process have environmental differences, resulting in model transplantation errors. These problems are particularly prominent in extreme working conditions such as high temperature and high pressure, multiphase flow, etc. Therefore, a high-precision electric calibration method and system for wedge flow meters is proposed. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides a high-precision electric calibration method and system for wedge flow meters to solve the problems raised in the background art.

[0006] To achieve the above purpose, the present application provides the following technical scheme: a high-precision electric calibration method for wedge flow meters, comprising the following steps:

[0007] Step 1, data acquisition and preprocessing:

[0008] Real-time acquisition of fluid temperature, pressure, density and wedge vibration signals through a multi-dimensional sensor array embedded in the pipeline, denoising of the original data using Kalman filtering algorithm, and construction of a multi-dimensional data set containing time domain features and frequency domain features;

[0009] Step 2, dynamic working condition identification:

[0010] Establish a working condition classification model based on support vector machines, input the processed data into the model for working condition identification, and trigger the calibration model update mechanism when significant changes in key working condition parameters are detected;

[0011] Step 3, adaptive model updating:

[0012] The recursive least square method is used to update the calibration model parameters online, an improved flow metering model containing temperature compensation term, pressure correction term and fluid compressibility compensation term is established, and the model updating period is dynamically adjusted according to the stability of the working condition;

[0013] Step four, multi-source data fusion calibration:

[0014] The wedge differential pressure signal and auxiliary sensor data are fused by using a data fusion algorithm, a weighted fusion strategy is constructed, the confidence weight of each sensor data is determined by an optimization algorithm, and the final calibration coefficient is generated;

[0015] Step five, closed loop verification and compensation;

[0016] The calibration result is input into the digital twin model for simulation verification, when the simulation error exceeds the set threshold, the error compensation mechanism is started, and the calibration process data and compensation result are stored in the distributed time series database, forming a closed loop control system of calibration, verification and compensation;

[0017] In the multi-dimensional sensor array, the thin film platinum resistance temperature sensor reflects the temperature by measuring the resistance change, the piezoresistive pressure sensor measures the pressure by using the piezoresistive effect, the vibrating string density meter measures the density according to the vibration frequency of the string, and the three-axis acceleration sensor measures the wedge vibration signal. The Kalman filter algorithm uses state equation and observation equation to iteratively update and remove noise from the original data;

[0018] Data acquisition and preprocessing obtain rich information through multi-dimensional sensor array, Kalman filter effectively denoises, provides high-quality data for subsequent processing, dynamic working condition recognition can adapt to changes in time, trigger calibration model update, ensure accuracy, adaptive model update adjusts dynamically according to working conditions, compensation term considers various factors, multi-source data fusion calibration improves data utilization, generates accurate calibration coefficient, closed loop verification and compensation ensures reliable calibration result, forms a complete closed loop control system, and improves the overall performance and stability of the flow meter.

[0019] Preferably, the multi-dimensional sensor array comprises:

[0020] a thin film platinum resistance temperature sensor;

[0021] a piezoresistive pressure sensor;

[0022] a vibrating string density meter and a three-axis acceleration sensor;

[0023] The thin film platinum resistance temperature sensor is used for high-precision measurement of fluid temperature, the piezoresistive pressure sensor is used for real-time monitoring of fluid pressure, the vibrating string density meter is used for measuring fluid density through vibration frequency change, and the three-axis acceleration sensor is used for capturing wedge vibration signal. Each sensor works cooperatively to ensure comprehensive and accurate data;

[0024] The Kalman filter and support vector machine algorithm are used to effectively remove noise and identify working condition changes, ensure dynamic adaptation of the calibration model, and further improve the measurement accuracy of the self-adaptive model updating and multi-source data fusion calibration technology. The closed-loop verification and compensation mechanism ensures the reliability of the calibration results. This method not only improves the measurement accuracy of the flow meter, but also enhances the adaptability and robustness of the system, providing an efficient and accurate solution for industrial measurement.

[0025] Preferably, the working condition classification model is trained using:

[0026] The kernel function is selected as a radial basis function.

[0027] A specific penalty factor and kernel parameter are set.

[0028] Typical working condition sample data is included.

[0029] First, sample data covering various typical working conditions is collected, including but not limited to data under different temperatures, pressures, fluid densities, and wedge vibration states. The radial basis function is selected as the kernel function, and the optimal penalty factor and kernel parameter are determined through cross-validation to ensure that the model can accurately classify under various working conditions. At the same time, batch gradient descent method is used to optimize the model parameters, improving the training efficiency and classification accuracy.

[0030] By selecting the radial basis function as the kernel function and setting the specific penalty factor and kernel parameter, the model can adapt to various complex working conditions, improve the classification accuracy, and ensure the wide applicability and robustness of the model. This method not only improves the accuracy of working condition recognition, but also provides a solid foundation for subsequent adaptive model updating and multi-source data fusion calibration, thereby realizing high-precision electrical calibration of the wedge flowmeter and ensuring the stability and reliability of flow measurement.

[0031] Preferably, the improved flow measurement model realizes dynamic compensation through the following methods:

[0032] Temperature compensation term: adjust the model parameters according to temperature changes.

[0033] Pressure correction term: real-time correction of pressure fluctuations.

[0034] Fluid compressibility compensation term: consider the influence of fluid density changes on measurement.

[0035] A temperature sensor feedback mechanism is embedded in the model to monitor the fluid temperature in real time, and the model parameters are dynamically adjusted according to a preset temperature-parameter mapping table; the pressure correction term is captured by a high-precision pressure sensor in real time to smooth the data by using a filtering algorithm, and then the pressure-related parameters in the model are corrected in real time; the fluid compressibility compensation term combines the density meter data to predict the fluid compressibility influence through the density change prediction model, and then compensates the measurement results;

[0036] The temperature compensation term ensures that the model parameters can be adaptively adjusted in different temperature environments, avoiding measurement errors caused by temperature changes; the pressure correction term effectively deals with the interference of fluid pressure fluctuations on measurement, realizing real-time correction of pressure fluctuations; the fluid compressibility compensation term fully considers the influence of fluid density change on measurement, so that the model can still maintain high precision under complex working conditions. The joint action of these measures makes the flow measurement results more reliable, meeting the demand for high-precision measurement in industrial sites.

[0037] Preferably, the data fusion algorithm adopts:

[0038] Establish a data fusion strategy;

[0039] Calculate the data conflict degree;

[0040] Execute the data fusion rule;

[0041] Generate a fusion confidence assessment;

[0042] First, based on the characteristics and importance of each sensor data, a data fusion strategy is developed, such as weighted average or Bayesian fusion; second, the correlation or difference between different sensor data is calculated to assess the degree of data conflict; then, the fusion rule is adjusted according to the conflict degree, such as directly fusing data with smaller conflicts and weighting adjusting data with larger conflicts; finally, the data confidence assessment after fusion is generated to ensure the reliability and accuracy of the fusion results;

[0043] By establishing a fusion strategy, this algorithm can comprehensively utilize multi-dimensional sensor data, reduce the influence of single sensor error on the overall measurement results, calculate the data conflict degree and execute the corresponding fusion rule, which can effectively handle the inconsistency between sensors, improve the rationality of data fusion, and finally generate a fusion confidence assessment, providing a quantitative basis for the accuracy of measurement results, which helps to discover and correct potential errors in time, thereby ensuring the stable operation and accurate measurement of the flowmeter under complex working conditions.

[0044] Preferably, the digital twin model construction includes:

[0045] Three-dimensional flow field simulation;

[0046] Structural finite element analysis

[0047] Sensor arrangement optimization

[0048] Real-time data-driven interface

[0049] In the three-dimensional flow field simulation, the computational fluid dynamics (CFD) technology is adopted, combined with the actual pipe size and fluid parameters, to establish an accurate three-dimensional flow field model; the structural finite element analysis is used to analyze the stress and strain of the wedge flowmeter structure, to ensure its mechanical performance; the sensor arrangement optimization selects the best arrangement scheme by simulating the influence of different sensor positions on the measurement accuracy; the real-time data-driven interface realizes the real-time interaction between the model and the actual measurement data through the API interface, to ensure that the model can dynamically reflect the actual working condition;

[0050] Through three-dimensional flow field simulation, the flow state of fluid in the pipe can be intuitively displayed, providing a theoretical basis for the design and optimization of the flowmeter; the structural finite element analysis ensures the structural stability of the flowmeter under complex working conditions, prolonging the service life; the sensor arrangement optimization improves the measurement accuracy and reduces the error; the real-time data-driven interface realizes the synchronous update of the model and the actual working condition, making the calibration result more accurate and reliable. This series of construction measures collectively improves the overall performance and practicality of the wedge flowmeter high-precision electric calibration method.

[0051] Preferably, the error compensation mechanism adopts:

[0052] The input layer nodes correspond to key parameters;

[0053] The hidden layer nodes perform feature extraction;

[0054] The output layer nodes generate compensation values;

[0055] The activation function selects a specific function:

[0056] In the error compensation mechanism, the input layer nodes will correspond to key parameters in the wedge flowmeter measurement, such as temperature, pressure, density, and vibration signals, etc. The hidden layer nodes perform nonlinear feature extraction on these parameters through a multi-layer neural network structure, to capture the complex relationships between data. The output layer nodes generate corresponding compensation values based on the extracted features, to correct the measurement error. The activation function selects a specific function such as ReLU (Rectified Linear Unit) or Sigmoid, to improve the nonlinear fitting ability of the neural network;

[0057] The error compensation mechanism realizes precise compensation of the measurement error of the wedge flowmeter through deep learning. By corresponding key parameters of the input layer nodes, it ensures that the compensation mechanism can comprehensively consider various factors affecting measurement. The feature extraction function of the hidden layer nodes enables the compensation mechanism to automatically learn and capture complex relationships between data, improving the accuracy and adaptability of compensation. The output layer nodes generate compensation values, which are directly used to correct measurement errors, improving the accuracy and reliability of measurement. In addition, the use of a specific activation function enhances the non-linear fitting ability of the neural network, enabling the compensation mechanism to better adapt to various complex working conditions and providing a strong guarantee for high-precision measurement of the wedge flowmeter.

[0058] Preferably, it also includes an abnormal condition handling mechanism:

[0059] When cavitation is detected, a specific correction module is started;

[0060] When the fluid is in a laminar flow state, switch to a specific calibration mode;

[0061] When two-phase flow is detected, activate a phase fraction compensation algorithm:

[0062] Cavitation detection can be achieved by analyzing the wedge vibration signal and pressure fluctuation characteristics. When the vibration frequency abnormally increases and the pressure fluctuation is severe, the correction module adjusts the flowmetering model parameters to eliminate the effects of cavitation. The laminar flow state switches to a specific calibration mode by identifying that the fluid flow rate is below a critical value, enabling a low flow rate dedicated calibration curve. Two-phase flow detection uses a densimeter and multiphase flow analysis algorithm. When the fluid density change exceeds a pre-set threshold, the phase fraction compensation algorithm is activated to dynamically adjust the flow calculation based on the gas-liquid ratio.

[0063] The abnormal condition handling mechanism significantly improves the adaptability and measurement accuracy of the wedge flowmeter in complex working conditions. By specifically handling special conditions such as cavitation, laminar flow, and two-phase flow, the system can automatically adjust the working mode, effectively avoiding measurement errors that may occur in traditional flowmeters under these conditions. This not only ensures the stable operation of the flowmeter in various extreme environments, but also ensures the accuracy and reliability of the measurement data, providing a solid data foundation for industrial production process control and improving overall production efficiency and safety.

[0064] Preferably, the distributed time series database uses:

[0065] High-efficiency data compression technology;

[0066] Fast query response mechanism;

[0067] Supports multiple query methods;

[0068] The efficient data compression technology of the distributed time series database can reduce the storage space occupation by using advanced compression algorithms (such as LZ4, Zstandard, etc.) to compress the stored data in real time. The fast query response mechanism can be realized by building indexes (such as B+ tree, hash index) and optimizing query paths. The support for multiple query methods can meet the query needs of different users by providing SQL-like query interface, RESTful API or custom query language.

[0069] The fast query response mechanism enables users to quickly obtain the required data, improving the real-time performance and response speed of the system. The support for multiple query methods greatly facilitates users of different backgrounds, whether they are technical personnel or non-technical personnel, can perform data retrieval through familiar query interfaces, enhancing the usability and flexibility of the system. These features collectively improve the overall performance and user experience of the system, providing solid technical support for data management and analysis in the flow meter calibration process.

[0070] The system for high-precision electrical calibration of wedge flow meters adopts the high-precision electrical calibration method of wedge flow meters described above, comprising:

[0071] Intelligent sensing layer: integrating a multi-dimensional sensor array;

[0072] Edge computing layer: deploying the high-precision electrical calibration method of wedge flow meters;

[0073] Cloud platform layer: conducting remote model training and parameter distribution;

[0074] Human-computer interaction layer: providing a calibration process visualization interface;

[0075] Thin-film platinum resistance temperature sensor for precise measurement of fluid temperature; piezoresistive pressure sensor for capturing pressure fluctuations; vibrating string density meter for measuring fluid density; three-axis acceleration sensor for monitoring wedge vibration. These sensors work together to perform denoising processing on the collected data through Kalman filtering algorithm, ensuring the accuracy and reliability of the data.

[0076] The system for high-precision electrical calibration of wedge flow meters integrates intelligent sensing layer, edge computing layer, cloud platform layer and human-computer interaction layer, realizing efficient and accurate flow measurement and calibration. The intelligent sensing layer uses a multi-dimensional sensor array to collect fluid parameters in real time, the edge computing layer deploys calibration algorithms for real-time processing, the cloud platform layer is responsible for remote model training and parameter distribution, and the human-computer interaction layer provides an intuitive visualization interface for operators to monitor and adjust. This system architecture not only improves calibration accuracy, but also enhances system flexibility and maintainability, ensuring stable operation under different working conditions.

[0077] Compared with the prior art, the wedge flowmeter high-precision electric calibration method and system provided by the present application has the following beneficial effects:

[0078] In the calibration process of the wedge flowmeter, the present application uses a multi-dimensional sensor array to collect fluid temperature, pressure, density and wedge vibration signals in real time, and uses Kalman filtering algorithm to denoise the original data to construct a multi-dimensional data set, achieving more comprehensive and accurate data collection and preprocessing, which provides a high-quality data basis for subsequent calibration, helps to improve measurement accuracy and reduce errors caused by inaccurate data.

[0079] Based on the support vector machine, a working condition classification model is established to recognize dynamic working conditions, and when significant changes in key working condition parameters are detected, a calibration model updating mechanism is triggered, so that the calibration process has dynamic self-adaptive ability and can respond to changes in working condition parameters in real time, effectively overcoming the defects of traditional calibration models lacking dynamic self-adaptive ability, and can adapt to complex and variable actual industrial scenes to ensure measurement accuracy.

[0080] The recursive least squares method is used to update the calibration model parameters online, an improved flowmetering model containing temperature compensation terms, pressure correction terms and fluid compressibility compensation terms is established, and the model updating period is dynamically adjusted according to the working condition stability, which enhances the representation ability of the calibration model for complex flow fields, avoids the limitations of a single sensor data source, and makes the calibration result more consistent with the actual working condition; the data fusion algorithm is used to fuse the wedge pressure difference signal and the auxiliary sensor data, a weighted fusion strategy is constructed, and the confidence weight of each sensor data is determined to generate the final calibration coefficient, further improving the accuracy of the calibration.

[0081] Finally, the calibration result is input into the digital twin model for simulation verification, and when the simulation error exceeds the threshold, an error compensation mechanism is started, and the calibration process data and compensation results are stored in a distributed time series database to form a closed-loop control system, effectively reducing the model transplantation error caused by the environmental difference between the calibration process and the actual measurement process, especially in extreme working conditions such as high temperature and high pressure, multiphase flow, etc. can significantly improve the measurement accuracy of the wedge flowmeter, ensure the stability of the production process and the product quality, and provide more reliable flow measurement support for industrial process control. BRIEF DESCRIPTION OF DRAWINGS

[0082] Figure 1 is the wedge flowmeter high-precision electric calibration method of the present application.

[0083] Figure 2 is the system schematic diagram of the wedge flowmeter high-precision electric calibration of the present application. DETAILED DESCRIPTION

[0084] The present application provides a technical solution, a wedge flowmeter high-precision electric calibration method, please refer to Figure 1, comprising the following steps:

[0085] Step one, data acquisition and pretreatment:

[0086] Real-time acquisition of fluid temperature, pressure, density and wedge vibration signals through the embedded multi-dimensional sensor array in the pipeline, denoising of the original data using Kalman filtering algorithm, and construction of a multi-dimensional data set containing time domain features and frequency domain features;

[0087] Step two, dynamic working condition identification:

[0088] Based on support vector machine to establish a working condition classification model, input the processed data into the model for working condition identification, when detecting significant changes in key working condition parameters, trigger the calibration model update mechanism;

[0089] Step three, adaptive model updating:

[0090] Recursive least squares method is used to update the calibration model parameters online, an improved flow metering model containing temperature compensation term, pressure correction term and fluid compressibility compensation term is established, and the model updating period is dynamically adjusted according to the working condition stability;

[0091] Step four, multi-source data fusion calibration:

[0092] Using data fusion algorithm to fuse wedge pressure difference signal and auxiliary sensor data, constructing a weighted fusion strategy, determining the confidence weight of each sensor data through optimization algorithm, and generating the final calibration coefficient;

[0093] Step five, closed-loop verification and compensation;

[0094] The calibration results are input into the digital twin model for simulation verification, when the simulation error exceeds the set threshold, the error compensation mechanism is started, and the calibration process data and compensation results are stored in the distributed time series database, forming a closed-loop control system of calibration, verification and compensation;

[0095] In the multi-dimensional sensor array, the thin film platinum resistance temperature sensor reflects the temperature by measuring the resistance change, the piezoresistive pressure sensor measures the pressure by using the piezoresistive effect, the vibrating string density meter measures the density according to the string vibration frequency, and the three-axis acceleration sensor measures the wedge vibration signal. Kalman filtering algorithm uses state equation and observation equation to iteratively update and remove noise from original data;

[0096] Data acquisition and preprocessing obtain rich information through a multi-dimensional sensor array, and Kalman filtering effectively removes noise to provide high-quality data for subsequent processing. Dynamic working condition recognition can adapt to changes in time, trigger calibration model updates, and ensure accuracy. Self-adaptive model updating dynamically adjusts according to working conditions, and the compensation term comprehensively considers multiple factors. Multi-source data fusion calibration improves data utilization, generates accurate calibration coefficients, and ensures reliable calibration results through closed-loop verification and compensation. This forms a complete closed-loop control system, improving the overall performance and stability of the flowmeter.

[0097] Please refer to Figure 1 , the multi-dimensional sensor array includes:

[0098] a thin-film platinum resistance temperature sensor;

[0099] a piezoresistive pressure sensor;

[0100] a vibrating string densimeter and a three-axis acceleration sensor;

[0101] The thin-film platinum resistance temperature sensor is used for high-precision measurement of fluid temperature, the piezoresistive pressure sensor is used for real-time monitoring of fluid pressure, the vibrating string densimeter is used for measuring fluid density through changes in vibration frequency, and the three-axis acceleration sensor is used to capture wedge vibration signals. Each sensor works together to ensure comprehensive and accurate data.

[0102] Kalman filtering and support vector machine algorithm are used to effectively remove noise and identify working condition changes, ensuring that the calibration model dynamically adapts. Self-adaptive model updating and multi-source data fusion calibration technology further improve measurement accuracy, and closed-loop verification and compensation mechanisms ensure the reliability of calibration results. This method not only improves the measurement accuracy of the flowmeter, but also enhances the adaptability and robustness of the system, providing an efficient and accurate solution for industrial measurement.

[0103] Please refer to Figure 1 , the working condition classification model training uses:

[0104] a radial basis function is selected as the kernel function;

[0105] a specific penalty factor and kernel parameter are set;

[0106] typical working condition sample data is included;

[0107] First, sample data covering various typical working conditions is collected, including but not limited to data under different temperatures, pressures, fluid densities, and wedge vibration states. A radial basis function is selected as the kernel function, and the optimal penalty factor and kernel parameter are determined through cross-validation to ensure that the model can accurately classify under various working conditions. At the same time, batch gradient descent method is used to optimize model parameters, improving training efficiency and classification accuracy.

[0108] By selecting the radial basis function as the kernel function and combining the specific penalty factor and the setting of the kernel parameter, the model can flexibly adapt to various complex working conditions, improve the classification accuracy, and ensure the wide applicability and robustness of the model. The training set containing a large number of typical working condition sample data not only improves the accuracy of working condition recognition, but also provides a solid foundation for subsequent adaptive model updating and multi-source data fusion calibration, thereby realizing high-precision electric calibration of the wedge flowmeter and ensuring the stability and reliability of flow measurement.

[0109] Please refer to Figure 1 The improved flow measurement model realizes dynamic compensation in the following ways:

[0110] Temperature compensation term: adjust model parameters according to temperature changes;

[0111] Pressure correction term: real-time correction of pressure fluctuations;

[0112] Fluid compressibility compensation term: consider the influence of fluid density changes on measurement;

[0113] In the model, a temperature sensor feedback mechanism is embedded to monitor the fluid temperature in real time, and the model parameters are dynamically adjusted according to the pre-set temperature-parameter mapping table. The pressure correction term uses a high-precision pressure sensor to capture pressure fluctuations in real time, and after smoothing the data using a filtering algorithm, it makes real-time corrections to the pressure-related parameters in the model. The fluid compressibility compensation term combines density meter data to predict the influence of fluid compressibility through a density change prediction model, and then compensates for the measurement results;

[0114] The temperature compensation term ensures that the model parameters can be adaptively adjusted in different temperature environments, avoiding measurement errors caused by temperature changes. The pressure correction term effectively deals with the interference of fluid pressure fluctuations on measurement, achieving real-time correction of pressure fluctuations. The fluid compressibility compensation term fully considers the influence of fluid density changes on measurement, so that the model can still maintain high precision in complex working conditions. These measures work together to make the flow measurement results more reliable, meeting the demand for high-precision measurement in industrial sites.

[0115] Please refer to Figure 1 The data fusion algorithm uses:

[0116] Establish data fusion strategy;

[0117] Calculate the degree of data conflict;

[0118] Execute data fusion rules;

[0119] Generate fusion confidence evaluation;

[0120] First, based on the characteristics and importance of each sensor's data, a data fusion strategy is formulated, such as weighted averaging or Bayesian fusion. Second, the degree of data conflict is assessed by calculating the correlation or difference between different sensor data. Then, the fusion rules are adjusted according to the degree of conflict, such as directly fusing data with less conflict and performing weighted adjustments on data with greater conflict. Finally, a confidence assessment of the fused data is generated to ensure the reliability and accuracy of the fusion results.

[0121] By establishing a fusion strategy, this algorithm can comprehensively utilize multi-dimensional sensor data, reduce the impact of single sensor errors on the overall measurement results, calculate the degree of data conflict and execute corresponding fusion rules, effectively handle data inconsistencies between sensors, improve the rationality of data fusion, and finally generate a fusion confidence assessment, which provides a quantitative basis for the accuracy of measurement results, helps to discover and correct potential errors in a timely manner, and thus ensures the stable operation and accurate measurement of the flow meter under complex working conditions.

[0122] Please see Figure 1 The construction of a digital twin model includes:

[0123] Three-dimensional flow field simulation;

[0124] Structural finite element analysis;

[0125] Sensor layout optimization;

[0126] Real-time data-driven interface;

[0127] In the three-dimensional flow field simulation, computational fluid dynamics (CFD) technology is used to establish an accurate three-dimensional flow field model by combining actual pipe dimensions and fluid parameters; structural finite element analysis uses finite element software to perform stress and strain analysis on the structure of the wedge flow meter to ensure its mechanical performance; sensor layout optimization selects the optimal layout scheme by simulating the impact of different sensor positions on measurement accuracy; and the real-time data-driven interface enables real-time interaction between the model and actual measurement data through an API interface, ensuring that the model can dynamically reflect the actual working conditions.

[0128] Three-dimensional flow field simulation can intuitively display the flow state of fluid in the pipeline, providing a theoretical basis for the design optimization of the flow meter; structural finite element analysis ensures the structural stability of the flow meter under complex working conditions and extends its service life; optimized sensor layout improves measurement accuracy and reduces errors; real-time data-driven interface enables synchronous updates between the model and actual working conditions, making the calibration results more accurate and reliable. These series of construction measures jointly improve the overall performance and practicality of the high-precision electrical calibration method for wedge flow meters.

[0129] Please see Figure 1 The error compensation mechanism adopts:

[0130] Input layer nodes correspond to key parameters;

[0131] Hidden layer nodes perform feature extraction;

[0132] Output layer nodes generate compensation values;

[0133] Activation function selects specific functions:

[0134] In the error compensation mechanism, the input layer nodes will correspond to the key parameters in the wedge flowmeter measurement, such as temperature, pressure, density and vibration signal, etc. The hidden layer nodes perform nonlinear feature extraction on these parameters through a multi-layer neural network structure to capture the complex relationship between the data. The output layer nodes generate corresponding compensation values according to the extracted features to correct the measurement error. The activation function selects ReLU (Rectified Linear Unit) or Sigmoid, etc. to improve the nonlinear fitting ability of the neural network;

[0135] The error compensation mechanism realizes accurate compensation of the wedge flowmeter measurement error through deep learning. By corresponding key parameters through input layer nodes, it ensures that the compensation mechanism can fully consider various factors affecting the measurement. The feature extraction function of the hidden layer nodes enables the compensation mechanism to automatically learn and capture the complex relationship between the data, improving the accuracy and adaptability of the compensation. The output layer nodes generate compensation values, which are directly used to correct the measurement error, improving the accuracy and reliability of the measurement. In addition, the use of specific activation functions enhances the nonlinear fitting ability of the neural network, making the compensation mechanism better adapt to various complex working conditions and providing a strong guarantee for high-precision measurement of the wedge flowmeter.

[0136] Please refer to Figure 1 It also includes an abnormal working condition processing mechanism:

[0137] When cavitation phenomenon is detected, start the specific correction module;

[0138] When the fluid is in laminar flow state, switch to specific calibration mode;

[0139] When two-phase flow is detected, activate phase fraction compensation algorithm:

[0140] Cavitation phenomenon detection can be achieved by analyzing the wedge vibration signal and pressure fluctuation characteristics. When the vibration frequency abnormally rises and the pressure fluctuation is severe, the correction module is triggered to adjust the flowmetering model parameters to eliminate the influence of cavitation. The laminar flow state switches to specific calibration mode by identifying that the fluid flow rate is below the critical value, and the low flow rate dedicated calibration curve is enabled. Two-phase flow detection uses a densimeter and multiphase flow analysis algorithm. When the fluid density change exceeds the preset threshold, the phase fraction compensation algorithm is activated to dynamically adjust the flow calculation according to the gas-liquid ratio;

[0141] The abnormal condition handling mechanism significantly improves the adaptability and measurement accuracy of the wedge flowmeter under complex conditions. Through special processing for cavitation, laminar flow, and two-phase flow, the system can automatically adjust the working mode, effectively avoiding the measurement errors that traditional flowmeters are prone to under these conditions. This not only ensures the stable operation of the flowmeter in various extreme environments, but also ensures the accuracy and reliability of the measurement data, providing a solid data foundation for industrial production process control and improving overall production efficiency and safety.

[0142] Please refer to Figure 1 , the distributed time series database uses:

[0143] efficient data compression technology;

[0144] fast query response mechanism;

[0145] supports multiple query methods;

[0146] The efficient data compression technology of the distributed time series database can compress the stored data in real time by using advanced compression algorithms (such as LZ4, Zstandard, etc.), reducing storage space occupation; the fast query response mechanism can be achieved by building indexes (such as B+ tree, hash index) and optimizing query paths; supporting multiple query methods can meet the query needs of different users by providing SQL-like query interface, RESTful API or custom query language;

[0147] The fast query response mechanism enables users to quickly obtain the required data, improving the real-time performance and response speed of the system; supporting multiple query methods greatly facilitates users of different backgrounds, whether they are technical personnel or non-technical personnel, can retrieve data through familiar query interfaces, enhancing the ease of use and flexibility of the system, which collectively improves the overall performance and user experience of the system, providing solid technical support for data management and analysis in the flowmeter calibration process.

[0148] The wedge flowmeter high-precision electrical calibration system uses the wedge flowmeter high-precision electrical calibration method described above, please refer to Figure 2 , including:

[0149] Intelligent sensing layer: integrates multi-dimensional sensor array;

[0150] Edge computing layer: deploy wedge flowmeter high-precision electrical calibration method;

[0151] Cloud platform layer: remote model training and parameter distribution;

[0152] Human-computer interaction layer: provides calibration process visualization interface;

[0153] The thin-film platinum resistance temperature sensor is used for precise measurement of fluid temperature; the piezoresistive pressure sensor is used for capturing pressure fluctuations; the vibrating string densimeter is used for measuring fluid density; the three-axis acceleration sensor is used for monitoring wedge vibration, which work together to ensure the accuracy and reliability of the data through Kalman filtering algorithm for denoising processing of collected data;

[0154] The system of high-precision electric calibration of wedge flowmeter realizes efficient and accurate flow measurement and calibration by integrating intelligent sensing layer, edge computing layer, cloud platform layer and human-computer interaction layer. The intelligent sensing layer uses multi-dimensional sensor array to collect fluid parameters in real time, the edge computing layer deploys calibration algorithms for real-time processing, the cloud platform layer is responsible for remote model training and parameter distribution, and the human-computer interaction layer provides intuitive visual interface for operators to monitor and adjust. This system architecture not only improves the calibration accuracy, but also enhances the flexibility and maintainability of the system, ensuring stable operation under different working conditions.

[0155] It should be noted that in this text, relational terms such as first and second are used only to distinguish one entity or operation from another, and do not necessarily require or imply any actual such relationship or order between these entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device.

[0156] Although embodiments of the present application have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and alterations can be made thereto without departing from the principles and spirit of the present application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A high-precision electrical calibration method for a wedge flowmeter, characterized in that, Includes the following steps: Step 1: Data Acquisition and Preprocessing The fluid temperature, pressure, density, and wedge vibration signals are collected in real time by a multi-dimensional sensor array embedded in the pipeline. The Kalman filter algorithm is used to denoise the raw data and construct a multidimensional dataset containing time-domain and frequency-domain features. Step 2: Dynamic Operating Condition Identification A working condition classification model is established based on support vector machine. The processed data is input into the model for working condition identification. When a significant change in key working condition parameters is detected, a calibration model update mechanism is triggered. Step 3: Adaptive Model Update The recursive least squares method is used to update the calibration model parameters online, and an improved flow metering model including temperature compensation, pressure correction and fluid compressibility compensation terms is established. The model update cycle is dynamically adjusted according to the stability of the operating conditions. Step 4: Multi-source data fusion calibration: A data fusion algorithm is used to fuse the wedge pressure differential signal and auxiliary sensor data to construct a weighted fusion strategy. The confidence weight of each sensor data is determined by an optimization algorithm to generate the final calibration coefficient. Step 5: Closed-loop verification and compensation; The calibration results are input into the digital twin model for simulation verification. When the simulation error exceeds the set threshold, the error compensation mechanism is activated. At the same time, the calibration process data and compensation results are stored in the distributed time series database, forming a closed-loop control system for calibration, verification and compensation. It also includes an abnormal operating condition handling mechanism: When cavitation is detected, a specific correction module is activated; When the fluid is in a laminar flow state, switch to a specific calibration mode; When two-phase flow is detected, the phase fraction compensation algorithm is activated; Cavitation detection analyzes wedge vibration signals and pressure fluctuation characteristics. When the vibration frequency rises abnormally and the pressure fluctuation is severe, the correction module is triggered to adjust the flow metering model parameters to eliminate the cavitation effect. The laminar flow state switching specific calibration mode identifies when the fluid velocity is below the critical value and activates a low-velocity dedicated calibration curve. Two-phase flow detection uses a densitometer and multiphase flow analysis algorithm. When the fluid density change is detected to exceed the preset threshold, the phase fraction compensation algorithm is activated to dynamically adjust the flow rate calculation according to the gas-liquid ratio.

2. The high-precision electrical calibration method for a wedge flowmeter according to claim 1, characterized in that, The multidimensional sensor array includes: Thin-film platinum resistance temperature sensor; Piezoresistive pressure sensor; Vibrating string densitometer and triaxial accelerometer.

3. The high-precision electrical calibration method for a wedge flowmeter according to claim 1, characterized in that, The working condition classification model is trained using: The kernel function is chosen to be a radial basis function; Set specific penalty factors and kernel parameters; It includes sample data for typical operating conditions.

4. The high-precision electrical calibration method for a wedge flowmeter according to claim 1, characterized in that, The improved flow metering model achieves dynamic compensation in the following way: Temperature compensation: Adjusts model parameters according to temperature changes; Pressure correction: Real-time correction for pressure fluctuations; Fluid compressibility compensation term: Considers the effect of fluid density changes on the measurement.

5. The high-precision electrical calibration method for a wedge flowmeter according to claim 1, characterized in that, The data fusion algorithm employs: Establish a data fusion strategy; Calculate the degree of data conflict; Execute data fusion rules; Generate a fusion confidence assessment.

6. The high-precision electrical calibration method for a wedge flowmeter according to claim 1, characterized in that, The construction of the digital twin model includes: Three-dimensional flow field simulation; Structural finite element analysis; Sensor layout optimization; Real-time data-driven interface.

7. The high-precision electrical calibration method for a wedge flowmeter according to claim 1, characterized in that, The error compensation mechanism adopts: Key parameters corresponding to input layer nodes; Feature extraction is performed on hidden layer nodes; Output layer nodes generate compensation values; The activation function is a specific function.

8. The high-precision electrical calibration method for a wedge flowmeter according to claim 1, characterized in that, The distributed time-series database adopts: High-efficiency data compression technology; Fast query response mechanism; Supports multiple query methods.

9. A system for high-precision electrical calibration of a wedge flow meter, employing a high-precision electrical calibration method for a wedge flow meter as described in any one of claims 1-8, characterized in that, include: Intelligent sensing layer: integrates a multi-dimensional sensor array; Edge computing layer: Deploying a high-precision electrical calibration method for wedge flow meters; Cloud platform layer: Performs remote model training and parameter distribution; Human-computer interaction layer: Provides a visual interface for the calibration process.

Images