Electromagnetic flowmeter online calibration system and acquisition and storage device

Abstract

The invention relates to the technical field of electromagnetic flowmeter calibration, and discloses an electromagnetic flowmeter online calibration system and an acquisition and storage device. The system comprises a signal acquisition module, a data processing module, an error analysis module, a parameter calibration module and a real-time monitoring module. The signal acquisition module captures an original signal waveform through a multi-channel sensor array and acquires flow signal data to generate a state set; the data processing module uses a wavelet packet transformation and adaptive filtering combined algorithm to reduce noise and extract features; the error analysis module constructs a dynamic time warping error matching model to obtain an error correction coefficient; the parameter calibration module adopts a fuzzy control and proportional integral differential composite strategy, and adjusts excitation current and electrode sensitivity according to an error correction coefficient; the real-time monitoring module establishes a rolling optimization model with time-varying constraints, and generates a calibration instruction in combination with signal stability and power consumption limitation. According to the system, online accurate calibration of the electromagnetic flowmeter is realized, and the measurement accuracy and the equipment operation stability are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electromagnetic flowmeter calibration, and particularly to an on-line calibration system for electromagnetic flowmeters and a collection and storage device. Background Art

[0002] As a key device in the field of industrial flow measurement, electromagnetic flowmeters are widely used in many industries such as petrochemical, water supply and drainage, and metallurgy due to their advantages of high measurement accuracy, wide range of measurable media, and no mechanical moving parts. However, in actual use, electromagnetic flowmeters are affected by various factors, resulting in a decrease in measurement accuracy, which seriously restricts the efficient operation and quality control of industrial production.

[0003] From the perspective of the measurement environment, the industrial site conditions are complex, and there are a large number of electromagnetic interference sources, such as the alternating electromagnetic fields generated by high-power motors, frequency converters and other equipment. These interferences will be superimposed on the measurement signal of the electromagnetic flowmeter, causing the distortion of the original signal waveform, and then affecting the accurate acquisition of the flow signal amplitude, frequency and phase difference data, resulting in an increase in measurement error. For example, in a chemical production workshop, due to the intensive operation of various electrical equipment, the measurement error of the electromagnetic flowmeter may reach several times that of normal conditions, affecting the precise control of the material flow, resulting in unstable product quality or waste of raw materials.

[0004] Long-term operation is also an important factor affecting the accuracy of electromagnetic flowmeters. As the use time increases, the sensor electrodes will gradually be corroded and scaled by the medium, changing the surface characteristics and sensitivity of the electrodes. At the same time, the performance of the excitation system will also decline due to component aging, resulting in unstable excitation current. These changes will cause the measurement characteristics of the electromagnetic flowmeter to drift and cannot accurately reflect the actual flow. According to statistics, among electromagnetic flowmeters that have been continuously operated for more than 1 year, about 30% of the equipment has a measurement error exceeding the allowable range, seriously affecting the reliability and economic benefits of the production process.

[0005] Traditional calibration methods have many limitations. Off-line calibration requires the electromagnetic flowmeter to be removed from the pipeline system and transported to a professional calibration laboratory for calibration. This process not only consumes a large amount of manpower, material resources and time, resulting in production interruption and increasing the production cost of enterprises, but also may damage the flowmeter during the disassembly and reinstallation process, further affecting its performance. For example, in a large water supply system, off-line calibration of the electromagnetic flowmeter may require shutting off the water supply for several hours or even several days, bringing great inconvenience to urban water supply and may also cause economic compensation problems.

[0006] Although online calibration technology has been gradually applied, most of the existing online calibration systems have problems such as low calibration accuracy and poor adaptability. Some systems use simple filtering algorithms to process measurement signals, which cannot effectively suppress complex interference, resulting in inaccurate calibration results. When faced with different working conditions, some systems cannot adjust the calibration strategy in a timely manner and are difficult to meet the diverse needs of industrial production. These problems limit the popularization and application of the online calibration technology for electromagnetic flowmeters, and there is an urgent need for a more advanced, efficient, and accurate online calibration system to solve the above problems. Summary of the Invention

[0007] The purpose of the present invention is to provide an online calibration system for electromagnetic flowmeters and a collection and storage device to solve the problems mentioned in the above background technology.

[0008] To achieve the above purpose, the present invention provides the following technical solutions: An online calibration system for electromagnetic flowmeters, the system includes: A signal acquisition module, which is used to synchronously capture the original signal waveform of the electromagnetic flowmeter through a multi-channel sensor array, and collect the amplitude, frequency, and phase difference data of the flow signal in real time to generate a flow state data set; A data processing module, which is used to perform noise suppression and feature extraction on the flow state data set by using a combined algorithm of wavelet packet transform and adaptive filtering to generate a preprocessed signal sequence; An error analysis module, which is used to construct an error matching model based on dynamic time warping, and generate an error correction coefficient set according to the deviation between the preset standard flow curve and the measured signal sequence; A parameter calibration module, which is used to design a composite regulation strategy based on fuzzy control and proportional integral differential, and dynamically adjust the excitation current and electrode sensitivity parameters of the electromagnetic flowmeter according to the error correction coefficient set to generate a calibrated parameter configuration; A real-time monitoring module, which is used to establish a rolling optimization model with time-varying constraints, and generate an online calibration instruction in combination with the signal stability threshold and the device power consumption limit.

[0009] Preferably, the combined algorithm of wavelet packet transform and adaptive filtering includes: Construct a multi-layer wavelet decomposition tree structure, and screen the optimal wavelet basis function through the energy entropy index; Design an order adjustment equation for the adaptive filter, and dynamically adjust the filtering bandwidth according to the signal-to-noise ratio; Adopt a sliding window mechanism to match the wavelet packet reconstruction and the filtering parameter update period, and output the denoised signal component.

[0010] Preferably, the construction of the error matching model includes: Set an elastic time window in the dynamic time warping stage, and calculate the local similarity between the measured signal and the standard curve through the covariance matrix; Introduce a weight distribution function in the error correction stage, and generate a correction coefficient priority list based on the classification of error amplitudes; Iteratively optimize the error matching path through the residual feedback mechanism to generate an error correction coefficient convergence report.

[0011] Preferably, the design of the composite adjustment strategy includes: Define the membership function of the fuzzy controller as a triangular distribution, and establish a fuzzy rule base for the excitation current deviation and the correction amount; Define the integral separation threshold of the proportional-integral-derivative controller, and construct the saturation constraint condition for the adjustment amount of the electrode sensitivity; Synchronously solve the fuzzy control and proportional-integral-derivative parameter combinations through the parallel computing framework to generate a piecewise continuous adjustment curve.

[0012] Preferably, the establishment of the rolling optimization model includes: Define the state variable as a combined vector of signal volatility, calibration response time, and power consumption level; Construct a model predictive control framework with a rolling time domain, and use the quadratic programming algorithm to solve the multi-objective optimization problem; Design the objective function as the normalized weighted sum of calibration accuracy and energy consumption cost, and output the optimal calibration action sequence.

[0013] Preferably, the system further includes: a signal traceability module, which is used to perform time-frequency characteristic analysis on the signals before and after calibration through Hilbert-Huang transform; input the analysis results into the historical database, and trigger a signal distortion warning signal through trend comparison.

[0014] Preferably, the generation of the flow state data set includes: synchronously collecting the instantaneous sampling rate, signal peak-to-peak value, and inter-channel cross-correlation data of the sensor array, and using the independent component analysis algorithm to optimize the multi-dimensional signal. Preferably, the system further includes: an equipment life evaluation module, which is used to establish an association model between the calibration times and sensor drift; use the genetic algorithm to dynamically optimize the calibration period and parameter adjustment amplitude, and generate the maximum calibration interval threshold.

[0015] Preferably, the generation of the online calibration instruction includes: Construct a state transition model with a Markov decision process, with signal stability and equipment life as boundary conditions; Use the dynamic programming algorithm to solve the multi-stage calibration strategy to generate a feasible solution set for the excitation current and electrode sensitivity; Evaluate the comprehensive utility value of the solution set through the entropy weight method, and output the real-time calibration operation queue.

[0016] Preferably, the present invention further includes an acquisition and storage device for an electromagnetic flowmeter online calibration system, and the device includes: A memory configured to store instructions; and A processor configured to call the instructions from the memory and capable of implementing the functions of the electromagnetic flowmeter online calibration system according to the above when executing the instructions.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: In terms of improving measurement accuracy, the signal acquisition module synchronously captures the original signal waveform through a multi-channel sensor array, collects rich flow signal data, and lays a foundation for accurate measurement. The data processing module adopts a combined algorithm of wavelet packet transform and adaptive filtering, which can effectively suppress noise and accurately extract features, improving the signal quality. The error analysis module constructs an error matching model based on dynamic time warping, which can accurately calculate the deviation from the standard flow curve and generate a reliable set of error correction coefficients, making the calibration more accurate.

[0018] In terms of intelligent and automatic calibration, the parameter calibration module designs a composite regulation strategy based on fuzzy control and proportional integral differential control, which can dynamically adjust the excitation current and electrode sensitivity parameters of the electromagnetic flowmeter according to the error correction coefficients. Fuzzy control can handle complex non-linear relationships, and proportional integral differential control ensures the stability and rapid response of the system. The combination of the two realizes intelligent and automatic parameter adjustment, reduces manual intervention, and improves the calibration efficiency. The real-time monitoring module establishes a rolling optimization model with time-varying constraints, generates online calibration instructions by combining the signal stability threshold and the device power consumption limit, and can trigger calibration operations in a timely and automatic manner according to the actual working conditions to ensure that the device is always in the best working state.

[0019] Regarding the operation stability and service life extension of the device, the signal traceability module analyzes the time-frequency characteristics of the signals before and after calibration through Hilbert-Huang transform, and triggers a signal distortion warning through trend comparison. It can timely detect potential problems, take measures in advance, avoid measurement errors and equipment failures caused by signal distortion, and ensure the operation stability of the device. The device life evaluation module establishes a correlation model between the calibration times and sensor drift, adopts a genetic algorithm to dynamically optimize the calibration period and parameter adjustment amplitude, generates a maximum calibration interval threshold, and a reasonable calibration strategy helps to slow down sensor drift, extend the service life of the device, and reduce the device maintenance cost.

[0020] In terms of energy consumption management, the rolling optimization model incorporates the calibration accuracy and energy consumption cost into the objective function, solves the multi-objective optimization problem through quadratic programming algorithm, and reduces the energy consumption of the device while ensuring the calibration accuracy.

[0021] In addition, the system also has good adaptability and versatility. Whether facing fluids with different medium characteristics or complex and changeable environmental conditions, through the collaborative work of each module, it can effectively perform calibration and meet the application requirements of various industrial scenarios. At the same time, the design of the acquisition and storage devices enables the stable realization of the system functions, ensuring the reliability and scalability of the entire online calibration system, and providing strong support for the efficient and stable operation of industrial production. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 is the working principle diagram of an online calibration system for an electromagnetic flowmeter according to the present invention; Figure 2 is the working flowchart for constructing an error matching model; Figure 3 is the working principle diagram for establishing a rolling optimization model; Figure 4 is the working principle diagram of a signal traceability module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0024] Please refer to Figures 1 - 4 , the present invention provides a technical solution: an online calibration system for an electromagnetic flowmeter, and its overall implementation solution is as follows: The system mainly consists of a signal acquisition module, a data processing module, an error analysis module, a parameter calibration module, and a real-time monitoring module. In actual operation, each module works collaboratively.

[0025] Signal acquisition module: This module synchronously captures the original signal waveform of the electromagnetic flowmeter through a multi-channel sensor array. During the capture process, the multi-channel sensor array obtains signals from different angles and positions to ensure the integrity and accuracy of the signals. At the same time, the amplitude, frequency, and phase difference data of the flow signal are collected in real time. For example, the amplitude reflects the intensity of the signal, the frequency reflects the speed of signal change, and the phase difference is used to analyze the time relationship between different signals. By collecting and processing these data, a flow state data set is generated, providing basic data for subsequent analysis and calibration.

[0026] Data processing module: The wavelet packet transform and adaptive filtering combined algorithm is used to process the flow state data set. This algorithm first constructs a multi-layer wavelet decomposition tree structure, and screens the optimal wavelet basis function through the energy entropy index to achieve effective decomposition and feature extraction of the signal. Then, the order adjustment equation of the adaptive filter is designed, and the filtering bandwidth is dynamically adjusted according to the signal-to-noise ratio, so as to effectively suppress noise in different noise environments. Finally, the sliding window mechanism is used to match the wavelet packet reconstruction and the filtering parameter update period, and the denoised signal component is output to obtain the preprocessed signal sequence, providing more accurate data for subsequent error analysis.

[0027] Error analysis module: Construct an error matching model based on dynamic time warping. Set an elastic time window in the dynamic time warping stage, and calculate the local similarity between the measured signal and the standard curve through the covariance matrix to measure the difference between the two. In the error correction stage, a weight distribution function is introduced, and a correction coefficient priority list is generated based on the error amplitude classification to determine the correction priority in different error situations. The error matching path is iteratively optimized through the residual feedback mechanism, and an error correction coefficient convergence report is generated to obtain the error correction coefficient set, providing a basis for subsequent parameter calibration.

[0028] Parameter calibration module: Design a composite adjustment strategy based on fuzzy control and proportional integral differential. Define the membership function of the fuzzy controller as a triangular distribution, establish a fuzzy rule base for the excitation current deviation and the correction amount, and perform fuzzy adjustment on the excitation current according to the error correction coefficient set. Define the integral separation threshold of the proportional integral differential controller, construct the saturation constraint condition of the electrode sensitivity adjustment amount, and precisely adjust the electrode sensitivity. The fuzzy control and proportional integral differential parameter combinations are synchronously solved through a parallel computing framework to generate a piecewise continuous adjustment curve, and the excitation current and electrode sensitivity parameters of the electromagnetic flowmeter are dynamically adjusted according to the error correction coefficient set to generate the calibrated parameter configuration.

[0029] Real-time monitoring module: Establish a rolling optimization model with time-varying constraints. Define the state variable as a combined vector of signal volatility, calibration response time, and power consumption level, and comprehensively consider multiple performance indicators of the system. Construct a model predictive control framework with a rolling time domain, and use the quadratic programming algorithm to solve the multi-objective optimization problem to find the optimal solution at different time points and conditions. Design the objective function as the normalized weighted sum of calibration accuracy and energy consumption cost, and balance the calibration accuracy and energy consumption by adjusting the weight. According to the above steps, the optimal calibration action sequence is output, and an online calibration instruction is generated in combination with the signal stability threshold and the device power consumption limit to achieve real-time calibration of the electromagnetic flowmeter.

[0030] The specific implementation manners of the present invention will be further described in detail below through 5 embodiments.

[0031] Example 1: This example elaborates in detail on the application of the combined algorithm of wavelet packet transform and adaptive filtering in the data processing module. In practical applications, the flow signals collected by electromagnetic flowmeters are often interfered by various noises, which affects the measurement accuracy. To solve this problem, this example uses the combined algorithm of wavelet packet transform and adaptive filtering to suppress noise and extract features from the flow state data set.

[0032] Construct a multi-layer wavelet decomposition tree structure, and select the optimal wavelet basis function through the energy entropy index. Let the number of layers of wavelet decomposition be , and in each layer of decomposition, there are multiple wavelet basis functions to choose from. The energy entropy index is used to measure the degree of energy distribution uniformity of the signal under different wavelet basis functions, and its calculation formula is: .

[0033] Among them, represents the proportion of the energy of the signal in the th frequency band to the total energy, and is the number of frequency bands. By calculating the energy entropy under different wavelet basis functions, select the wavelet basis function with the minimum energy entropy as the optimal wavelet basis function, so that the energy of the signal can be better concentrated in a few frequency bands, which is convenient for subsequent processing.

[0034] Design the order adjustment equation of the adaptive filter, and dynamically adjust the filter bandwidth according to the signal-to-noise ratio of the signal. Let the signal-to-noise ratio of the signal be , and the order of the adaptive filter be , and its adjustment equation is:

[0035] Among them, is the initial order, and is the adjustment coefficient. When the signal-to-noise ratio is relatively high, it indicates that the signal quality is good. At this time, the order of the filter can be appropriately reduced, and the filter bandwidth can be reduced to avoid losing signal features due to over-filtering; when the signal-to-noise ratio is relatively low, increase the order of the filter , and increase the filter bandwidth to better suppress noise.

[0036] Adopt a sliding window mechanism to match the wavelet packet reconstruction and the filter parameter update period. The size of the sliding window is set to . In each window, first perform wavelet packet reconstruction to obtain a preliminarily denoised signal. Then, according to the characteristics of the signal in this window, update the parameters of the adaptive filter. By continuously sliding the window, continuously process the signal, output the denoised signal component, and generate a preprocessed signal sequence. In this way, the processing parameters can be adjusted according to the real-time changes of the signal at different time points, improving the adaptability and processing effect of the algorithm.

[0037] Example 2: This example focuses on the construction of the error matching model in the error analysis module. In the calibration work of electromagnetic flowmeters, accurately measuring the deviation between the measured signal and the standard flow curve is a crucial part of the calibration.

[0038] In the dynamic time warping stage, an elastic time window is set to calculate the local similarity between the measured signal and the standard curve. Suppose the measured signal sequence is , and the standard curve sequence is . Set the size of the elastic time window to be . For each point in the measured signal, when searching for the matching point in the standard curve, the search range is limited to . When calculating the similarity between the two, a simple distance formula is used for measurement, that is . By traversing all the points within the window, find the with the smallest distance from . This minimum distance represents the local similarity of the current point. For example, when , and takes values of within the window, calculate the distances , , respectively. Then the minimum distance is 1, and the matching points found at this time are or (the unique matching point can be further determined according to other rules).

[0039] Enter the error correction stage, introduce a weight assignment function, classify according to different error amplitudes, and then generate a list of correction coefficient priorities. Set the error amplitude to be , and the weight assignment function is , specifically defined as: , where is the error amplitude threshold set according to actual calibration requirements and experience. When the error amplitude is greater than or equal to , it indicates that this error has a greater impact on the measurement result. At this time, assign a weight of , meaning that the correction coefficient corresponding to this error has a higher priority and needs to be corrected first; while when the error amplitude is less than , the weight , indicating that this error is relatively small and can be processed later. In this way, all error points are sorted according to the size of the error amplitude to generate a list of correction coefficient priorities.

[0040] Finally, with the help of the residual feedback mechanism, the error matching path is iteratively optimized to generate a convergence report of the error correction coefficient. After each correction, the residual is calculated , and the residual is calculated as , where is the measured signal value after correction. The residual is fed back to the next matching process, and the matching path is adjusted according to the magnitude and trend of the residual. For example, if the residual gradually decreases after multiple iterations, it indicates that the current matching path adjustment direction is correct; if the residual increases, the matching strategy needs to be re-examined. This iterative process is continuously repeated until the residual is less than a pre-set threshold. At this time, it is considered that the error matching path has converged, and then a convergence report of the error correction coefficient is generated to obtain an accurate set of error correction coefficients, providing a reliable basis for subsequent parameter calibration.

[0041] Embodiment 3: This embodiment mainly focuses on the design and application of the composite adjustment strategy in the parameter calibration module. To achieve precise adjustment of the excitation current and electrode sensitivity parameters of the electromagnetic flowmeter, a composite adjustment strategy based on fuzzy control and proportional-integral-derivative is adopted.

[0042] The membership function of the fuzzy controller is defined as a triangular distribution, and a fuzzy rule base for the excitation current deviation and correction amount is constructed. Let the excitation current deviation be , and its universe of discourse is set as , and the fuzzy subsets include , representing negative large (NB), negative medium (NM), negative small (NS), zero (ZE), positive small (PS), positive medium (PM), and positive large (PB) respectively. Taking the positive small (PS) fuzzy subset as an example, its membership function is: .

[0043] Among them, and are parameters determined according to the actual situation and experimental data. determines the position where the membership function starts to rise from 0, and determines the position where the membership function drops to 0. Based on a large number of experiments and professional experience, a fuzzy rule base for the excitation current deviation and correction amount is established. For example, when it is detected that the excitation current deviation is negative large (NB), according to the fuzzy rule base, the correction amount is determined to be positive large (PB) to achieve the preliminary adjustment of the excitation current.

[0044] Define the integral separation threshold of the proportional-integral-derivative controller, and simultaneously construct the saturation constraint condition of the electrode sensitivity adjustment amount. Let the electrode sensitivity be , the adjustment amount be , the input of the proportional-integral-derivative controller is the error , set the integral separation threshold as . When the absolute value of the error is greater than the integral separation threshold , to prevent the occurrence of integral saturation phenomenon, temporarily cancel the integral action; when , the integral link works normally. To ensure the stable operation of the system, the saturation constraint condition of the electrode sensitivity adjustment amount is constructed as: , where and are the minimum and maximum values of the electrode sensitivity adjustment amount set according to the device performance and safety requirements respectively. This avoids the situation where the system becomes unstable or even the device is damaged due to too large or too small adjustment amount.

[0045] Use the parallel computing framework to synchronously solve the fuzzy control and proportional-integral-derivative parameter combinations, and then generate a piecewise continuous adjustment curve. With the powerful computing power of the parallel computing framework, it can calculate the parameter combinations of fuzzy control and proportional-integral-derivative control simultaneously, greatly improving the computing efficiency. Under different error conditions, according to the calculated parameter combinations, generate the corresponding adjustment curve segments. These curve segments are connected to form a piecewise continuous adjustment curve. Through this curve, according to the error correction coefficient set, dynamically and accurately adjust the excitation current and electrode sensitivity parameters of the electromagnetic flowmeter, and finally generate the calibrated parameter configuration to complete the high-precision calibration of the electromagnetic flowmeter.

[0046] Example 4: This example details the establishment and application of the rolling optimization model in the real-time monitoring module. In order to ensure the calibration accuracy while taking into account the power consumption and response time of the device, this example establishes a rolling optimization model with time-varying constraints.

[0047] Define the state variable as a combined vector of signal volatility, calibration response time, and power consumption level. Let the signal volatility be , the calibration response time be , the power consumption level be , and the state variable vector . The signal volatility reflects the stability of the signal and is obtained by calculating the standard deviation of the signal over a period of time; the calibration response time represents the time from issuing the calibration command to completing the calibration; the power consumption level reflects the energy consumption of the device during the calibration process.

[0048] Construct a model predictive control framework with a rolling horizon and use a quadratic programming algorithm to solve the multi-objective optimization problem. Let the rolling horizon be , and at each time step , based on the current state variables predict the states in the future time steps. The objective function of the multi-objective optimization problem is the normalized weighted sum of the calibration accuracy and the energy consumption cost , that is: .

[0049] Among them, and are weight coefficients, and . The calibration accuracy can be measured by the error between the measured flow rate and the standard flow rate, and the energy consumption cost is related to the power consumption level . Use the quadratic programming algorithm to solve this multi-objective optimization problem to obtain the optimal control input at the current time step.

[0050] Design the objective function as the normalized weighted sum of the calibration accuracy and the energy consumption cost, and output the optimal calibration action sequence. By adjusting the weight coefficients and , the calibration accuracy and the energy consumption can be balanced according to the actual requirements. For example, when a higher calibration accuracy is required, increase ; when more sensitive to energy consumption, increase . According to the obtained optimal control input, output the optimal calibration action sequence, and generate an online calibration instruction by combining the signal stability threshold and the device power consumption limit to achieve real-time calibration of the electromagnetic flowmeter.

[0051] Example 5: This example covers the relevant content of the signal traceability module, the device life evaluation module, and the generation of online calibration instructions.

[0052] The signal traceability module performs time-frequency characteristic analysis on the signals before and after calibration through Hilbert-Huang transform. Let the signal before calibration be , and the signal after calibration be . The Hilbert-Huang transform first performs empirical mode decomposition on the signal, decomposing the signal into multiple intrinsic mode functions , that is: .

[0053] Then perform Hilbert transform on each intrinsic mode function to obtain its instantaneous frequency and amplitude . By analyzing the changes in the time-frequency characteristics of the signals before and after calibration, such as changes in frequency components and amplitude adjustments, the calibration effect can be judged. The analysis results are input into the historical database, and the signal distortion warning signal is triggered through trend comparison. When abnormal changes are found in the time-frequency characteristics of the calibrated signal compared with the historical data, a warning is triggered, indicating that there may be a signal distortion problem.

[0054] The equipment life assessment module establishes the correlation model between the calibration times and sensor drift . Let the correlation model be . Through the analysis and fitting of a large amount of historical data, the specific form of the function is determined. The genetic algorithm is used to dynamically optimize the calibration period and the parameter adjustment amplitude to generate the maximum calibration interval threshold. The genetic algorithm simulates the process of natural selection and genetic variation to search for the optimal calibration period and the parameter adjustment amplitude within a certain range. The objective function can be set as a function that comprehensively considers factors such as equipment life, calibration cost, and measurement accuracy, such as: .

[0055] Among them, represents the remaining life of the equipment, represents the calibration cost, represents the measurement accuracy, is the weight coefficient. Through the iterative optimization of the genetic algorithm, the optimal calibration period and the parameter adjustment amplitude are obtained, and then the maximum calibration interval threshold is generated.

[0056] The generation process of the online calibration instruction is as follows: Construct a state transition model with a Markov decision process, with signal stability and equipment life as boundary conditions. Let the state of the system be , the action set be , and the state transition probability be , where is the new state of the system after executing the action . Under the constraints of signal stability and equipment life, the rules of state transition are determined. The dynamic programming algorithm is used to solve the multi-stage calibration strategy to generate a feasible solution set of excitation current and electrode sensitivity. The dynamic programming algorithm decomposes the multi-stage decision problem into a series of sub-problems and gradually solves to obtain the optimal strategy. The comprehensive utility value of the solution set is evaluated by the entropy weight method, and the real-time calibration operation queue is output. The entropy weight method determines the weights of each index according to the variation degree of each scheme index, thereby evaluating the comprehensive utility value of different feasible solutions, and selects the scheme with the highest comprehensive utility value as the real-time calibration operation queue to achieve efficient and accurate online calibration of the electromagnetic flowmeter.

[0057] Although embodiments of the present invention have been shown and described, those of ordinary skill in the art will appreciate that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An electromagnetic flowmeter online calibration system, characterized in that: The system comprises: The signal acquisition module is used to synchronously capture the original signal waveform of the electromagnetic flowmeter through a multi-channel sensor array, and collect the amplitude, frequency and phase difference data of the flow signal in real time to generate a flow state data set; A data processing module, used for performing noise suppression and feature extraction on the flow state data set by adopting a wavelet packet transform and adaptive filtering joint algorithm to generate a preprocessing signal sequence; The error analysis module is used to construct an error matching model based on dynamic time warping and generate a set of error correction coefficients according to the deviation between the preset standard flow curve and the measured signal sequence; A parameter calibration module is used to design a composite regulation strategy based on fuzzy control and proportional integral differential, dynamically adjust the excitation current and electrode sensitivity parameters of the electromagnetic flowmeter according to the error correction coefficient set, and generate a calibrated parameter configuration; The real-time monitoring module is used to establish a rolling optimization model with time-varying constraints and generate online calibration instructions based on the signal stability threshold and device power consumption limit.

2. The electromagnetic flowmeter online calibration system according to claim 1, characterized in that: The wavelet packet transform and adaptive filtering joint algorithm includes: Construct a multi-layer wavelet decomposition tree structure and select the optimal wavelet basis function through the energy entropy index; Design the order adjustment equation of the adaptive filter and dynamically adjust the filter bandwidth according to the signal-to-noise ratio; The sliding window mechanism is used to match the wavelet packet reconstruction and filtering parameter update period, and the denoised signal components are output.

3. The electromagnetic flowmeter online calibration system according to claim 1, characterized in that: The construction of the error matching model includes: In the dynamic time warping stage, an elastic time window is set, and the local similarity between the measured signal and the standard curve is calculated through the covariance matrix; In the error correction stage, a weight allocation function is introduced to generate a priority list of correction coefficients based on the error amplitude classification; The error matching path is iteratively optimized through the residual feedback mechanism, and an error correction coefficient convergence report is generated.

4. The electromagnetic flowmeter online calibration system according to claim 3, characterized in that: The design of the composite regulation strategy includes: The membership function of the fuzzy controller is defined as a triangular distribution, and a fuzzy rule base for excitation current deviation and correction is established; Define the integral separation threshold of the proportional integral derivative controller and construct the saturation constraint condition of the electrode sensitivity adjustment amount; The fuzzy control and proportional-integral-differential parameter combinations are solved synchronously through a parallel computing framework to generate a piecewise continuous regulation curve.

5. The electromagnetic flowmeter online calibration system according to claim 1, characterized in that: The establishment of the rolling optimization model includes: The state variable is defined as a combination vector of signal fluctuation rate, calibration response time and power consumption level; Construct a model predictive control framework with a rolling horizon and use a quadratic programming algorithm to solve multi-objective optimization problems; The objective function is designed as the normalized weighted sum of calibration accuracy and energy consumption cost, and the optimal calibration action sequence is output.

6. The electromagnetic flowmeter online calibration system according to claim 1, characterized in that: The system also includes: a signal tracing module, which is used to analyze the time-frequency characteristics of the signals before and after calibration through Hilbert-Huang transform; input the analysis results into a historical database, and trigger a signal distortion warning signal through trend comparison.

7. The electromagnetic flowmeter online calibration system according to claim 1, characterized in that: The generation of the flow state data set includes: synchronously collecting the instantaneous sampling rate, signal peak-to-peak value and channel-to-channel cross-correlation data of the sensor array, using an independent component analysis algorithm to perform blind source separation on multidimensional signals, and constructing a characteristic signal subspace.

8. The electromagnetic flowmeter online calibration system according to claim 1, characterized in that: The system also includes: an equipment life assessment module for establishing a correlation model between the number of calibrations and sensor drift; and a genetic algorithm for dynamically optimizing the calibration cycle and parameter adjustment range to generate a maximum calibration interval threshold.

9. The electromagnetic flowmeter online calibration system according to claim 1, characterized in that: The generation of the online calibration instruction includes: Construct a state transition model with Markov decision process, with signal stability and equipment life as boundary conditions; A dynamic programming algorithm is used to solve the multi-stage calibration strategy and generate feasible solutions of excitation current and electrode sensitivity. The comprehensive utility value of the solution set is evaluated by the entropy weight method, and a real-time calibration operation queue is output.

10. A collection and storage device for an electromagnetic flowmeter online calibration system, characterized in that: The device comprises: A memory configured to store instructions; and a processor configured to call the instructions from the memory and implement the functions of the electromagnetic flowmeter online calibration system according to any one of claims 1 to 9 when executing the instructions.

Images