High-efficiency and high-precision compensation system and method for intelligent mutual inductor

Abstract

The invention discloses an efficient and high-precision compensation system and method of an intelligent mutual inductor, and particularly relates to the technical field of mutual inductors. The method comprises the following steps: acquiring magnetic flux density and magnetic field intensity data of a mutual inductor under different current and voltage change rates, generating a hysteresis loop diagram, extracting hysteresis loop width, residual magnetism and coercive force characteristics, and establishing a hysteresis effect error model; the harmonic distortion is identified in combination with spectrum analysis, the harmonic distortion degree in a high-frequency environment is predicted, the hysteresis of the hysteresis effect and the harmonic distortion are comprehensively analyzed, and the accuracy of a compensation algorithm for processing nonlinear errors is evaluated; and finally, the compensation signals are divided into high-accuracy signals and low-accuracy signals, the high-accuracy signals are directly output, and the low-accuracy signals are output after error correction, so that the problem that a conventional compensation algorithm cannot correct non-linear errors is effectively solved, and the measurement precision of the mutual inductor in an extreme environment and the operation stability of a system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data visualization, and in particular to a high-efficiency and high-precision compensation system and method for an intelligent mutual inductor. Background Art

[0002] The efficient and high-precision compensation of intelligent transformers refers to the dynamic adjustment and compensation of the transformer errors under different working conditions through intelligent algorithms and digital signal processing technology, thereby improving the measurement accuracy of the transformer. Since the transformer is affected by various factors such as environment, load and temperature during operation, which may cause the output signal to deviate from the actual signal, the compensation function of the intelligent transformer can effectively correct these errors and ensure the accuracy and reliability of the data.

[0003] In addition, efficient and high-precision compensation not only improves the measurement accuracy of the transformer, but also improves its working efficiency. Through real-time monitoring and adjustment, the intelligent transformer can reduce energy loss and resource waste without increasing hardware complexity, thereby achieving more energy-saving operation. This compensation technology is widely used in power systems to help ensure the stable transmission and distribution of electrical energy.

[0004] The prior art has the following deficiencies:

[0005] In extreme electromagnetic environments or high-frequency, high-voltage operating scenarios, transformers may produce serious nonlinear errors that cannot be effectively corrected by conventional compensation algorithms. For example, the core of the transformer may saturate under extreme conditions, resulting in an aggravated nonlinear response of the induced signal. Intelligent algorithms are usually based on linear models or assume that the errors of the transformer are predictable, but nonlinear errors are often difficult to model accurately. If the system fails to take these nonlinear characteristics into account, the compensation algorithm may make incorrect adjustments, causing the transformer output signal to become more distorted. And long-term error accumulation may even cause equipment failure or unstable operation of the power system, especially in environments with high precision requirements, such as power dispatch centers or automation systems for protection equipment. Summary of the invention

[0006] The object of the present invention is to provide a high-efficiency and high-precision compensation system and method for an intelligent mutual inductor, so as to solve the deficiencies in the background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: an efficient and high-precision compensation method for an intelligent mutual inductor, comprising the following steps:

[0008] S1: Under different current and voltage change rates, the magnetic flux density and magnetic field strength data of the transformer are collected to generate a hysteresis loop diagram;

[0009] S2: According to the hysteresis loop diagram, the hysteresis loop width characteristics and the residual magnetism and coercive force characteristics in the hysteresis loop are extracted respectively, and the hysteresis effect error model of the mutual inductor is established. The hysteresis of the hysteresis effect is determined according to the output results of the model;

[0010] S3: Use digital signal processing to perform spectrum analysis on the output signal of the transformer, identify the harmonic distortion of different frequency components in the signal, and predict the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment;

[0011] S4: Comprehensively analyze the hysteresis of the hysteresis effect and the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment to evaluate the signal accuracy of the compensation algorithm when dealing with the nonlinear error of the transformer;

[0012] S5: The signal accuracy of the compensation algorithm in response to the nonlinear error of the transformer is divided into different levels, that is, high-accuracy signals and low-accuracy signals. The high-accuracy signal is directly used as the final output signal of the transformer, and the low-accuracy signal is output after error correction.

[0013] Preferably, in S2, according to the hysteresis loop diagram, the hysteresis loop width characteristics and the residual magnetism and coercive force characteristics in the hysteresis loop are respectively extracted, a hysteresis effect error model of the mutual inductor is established, and the hysteresis of the hysteresis effect is determined according to the output result of the model, specifically:

[0014] The hysteresis loop width indicates the degree of hysteresis between the magnetic field intensity H and the magnetic flux density B during the magnetization and demagnetization process of the transformer, and is calculated by measuring the difference in magnetic field intensity when the magnetic flux density is zero;

[0015] In the hysteresis loop diagram, find the point where the magnetic flux density B = 0. When B = 0, record the positive and negative magnetic field strengths, record them as Hpositive and Hnegative respectively, and calculate the expression: Wh = Hpositive-Hnegative; where: Wh is the width of the hysteresis loop, reflecting the hysteresis in the magnetization and demagnetization process, Hpositive is the positive magnetic field strength when the magnetic flux density is zero, and Hnegative is the negative magnetic field strength when the magnetic flux density is zero;

[0016] Remanence is the value of magnetic flux density B when magnetic field strength H=0, indicating the magnetic flux retained in the core of the transformer after demagnetization. In the hysteresis loop diagram, find the point where magnetic field strength H=0, record the magnetic flux density value at this time, recorded as Br, and calculate the expression: Br=B(H=0); where Br is remanence, indicating the magnetic flux density in the magnetization loop when the magnetic field strength is zero;

[0017] Coercive force is the reverse magnetic field strength required to make the magnetic flux density B = 0, which indicates the magnetic field strength required by the transformer core to eliminate residual magnetism. In the hysteresis loop diagram, find the point where the magnetic flux density B = 0, record the corresponding reverse magnetic field strength, recorded as Hc, and calculate the expression: Hc = |H(B = 0)|; where Hc is the coercive force, which indicates the reverse magnetic field strength required to return the magnetic flux density to zero;

[0018] The hysteresis effect error model reflects the nonlinear error of the hysteresis effect in the operation of the transformer. The hysteresis effect error model is established through the hysteresis loop width Wh, remanence Br, and coercive force Hc. The hysteresis effect error is described as the hysteresis relationship between the input magnetic field intensity and the actual output magnetic flux density. Under a given magnetic field intensity H, the error ε caused by the hysteresis effect is mag (H) The expression is: mag (H) = α·Wh+β·Br+γ·Hc; wherein α, β, and γ are unknown coefficients.

[0019] Preferably, in S3, digital signal processing is used to perform spectrum analysis on the output signal of the transformer, to identify the harmonic distortion of different frequency components in the signal, and to predict the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment, specifically:

[0020] The output signal of the transformer is collected in real time. Before spectrum analysis, the DC component is eliminated by removing the mean value of the signal. The expression is: Where x(t) is the original signal, N is the number of signal samples, and a window function is applied to the signal to enhance the resolution of the spectrum within a limited sampling length. The expression is: windowed =x(t)·w(t); where w(t) is a window function; the preprocessed time domain signal x(t) is subjected to a fast Fourier transform operation to obtain a frequency domain signal X(f), which is expressed as follows: X(k) is the kth discrete frequency component in the frequency domain, N is the total number of sampling points, k represents different frequency components, x(n) is the nth sampling point in the time domain signal, is the kernel function of Fourier transform. After obtaining the frequency domain data through fast Fourier transform operation, a spectrum diagram is drawn, with the horizontal axis being the frequency and the vertical axis being the signal amplitude. In the spectrum diagram, the harmonic signal will appear as a peak value at an integer multiple of the fundamental frequency f0;

[0021] Through spectrum analysis, the harmonic distortion in the signal is identified and quantified, which is the ratio of the sum of the squares of all harmonic components in the signal to the square of the fundamental component. The expression is: Where THD is the total harmonic distortion, V1 is the amplitude of the fundamental frequency, and V8, V9, ... are the amplitudes of each harmonic.

[0022] Preferably, the harmonic distortion degree of the output signal of the transformer in a high-frequency environment is predicted to generate a harmonic distortion anomaly index, and the method for obtaining the harmonic distortion anomaly index is:

[0023] Select the wavelet basis function, perform multi-scale wavelet decomposition on the harmonic distortion data, obtain the detail coefficients and approximate coefficients at different scales, use the detail coefficients and approximate coefficients to reconstruct the signal, and obtain the reconstructed signal at different scales. The specific calculation expression is: Among them, AX j％,- and DX j,- are the reconstructed approximate coefficient and detail coefficient respectively, and BX(t) is the reconstructed signal; the harmonic distortion anomaly index is calculated according to the reconstructed signal, and the variation of the output signal of the transformer is obtained by calculating the energy change of the detail coefficient at different scales, and the energy of the detail coefficient at each scale is calculated. The specific calculation expression is: E j =∑ - AD j,- A 8 ; Among them, E j is the energy of the jth layer. The harmonic distortion anomaly index is calculated according to the energy changes at different scales. The specific calculation expression is: Where QV is the harmonic distortion anomaly index.

[0024] Preferably, in S4, the hysteresis of the hysteresis effect and the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment are comprehensively analyzed, specifically:

[0025] The hysteresis effect and the harmonic distortion anomaly index are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model uses each group of first eigenvectors to predict the signal accuracy value label of the compensation algorithm when dealing with the transformer nonlinear error as the prediction target, and takes minimizing the sum of the prediction errors of the signal accuracy value labels of all compensation algorithms when dealing with the transformer nonlinear error as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The signal accuracy value of the compensation algorithm when dealing with the transformer nonlinear error is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0026] Preferably, in S5, the signal accuracy of the compensation algorithm in dealing with the nonlinear error of the mutual inductor is divided into different levels, which are divided into high-accuracy signals and low-accuracy signals, specifically:

[0027] The obtained signal accuracy value of the compensation algorithm in response to the nonlinear error of the mutual inductor is compared with the preset reference threshold of the signal accuracy value. If the signal accuracy value of the compensation algorithm in response to the nonlinear error of the mutual inductor is greater than or equal to the reference threshold of the signal accuracy value, it means that the signal accuracy of the compensation algorithm in response to the nonlinear error of the mutual inductor is high. At this time, no warning signal is generated, and the signal of the compensation algorithm in response to the nonlinear error of the mutual inductor is classified as a high-accuracy signal, which is directly used as the final output signal of the mutual inductor;

[0028] If the signal accuracy value when the compensation algorithm responds to the nonlinear error of the transformer is less than the reference threshold value of the signal accuracy value, it means that the signal accuracy when the compensation algorithm responds to the nonlinear error of the transformer is low. At this time, a warning signal is generated, and the signal when the compensation algorithm responds to the nonlinear error of the transformer is classified as a low-accuracy signal, which is output after error correction.

[0029] The present invention also provides a high-efficiency and high-precision compensation system for an intelligent mutual inductor, including a data acquisition module, a hysteresis effect modeling module, a spectrum analysis and harmonic detection module, a comprehensive error analysis and evaluation module, and a signal classification and output module;

[0030] Data acquisition module: collects the flux density and magnetic field strength data of the transformer at different current and voltage change rates to generate a hysteresis loop diagram;

[0031] Hysteresis effect modeling module: According to the hysteresis loop diagram, the hysteresis loop width characteristics, residual magnetism and coercive force characteristics in the hysteresis loop are extracted respectively, and the hysteresis effect error model of the mutual inductor is established. The hysteresis of the hysteresis effect is determined according to the output results of the model;

[0032] Spectrum analysis and harmonic detection module: Use digital signal processing to perform spectrum analysis on the output signal of the transformer, identify the harmonic distortion of different frequency components in the signal, and predict the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment;

[0033] Comprehensive error analysis and evaluation module: Comprehensively analyze the hysteresis of the hysteresis effect and the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment, and evaluate the signal accuracy of the compensation algorithm when dealing with the nonlinear error of the transformer;

[0034] Signal classification and output module: The signal accuracy when the compensation algorithm responds to the nonlinear error of the transformer is divided into different levels, that is, high-accuracy signals and low-accuracy signals. The high-accuracy signal is directly used as the final output signal of the transformer, and the low-accuracy signal is output after error correction.

[0035] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0036] 1. The present invention collects the magnetic flux density and magnetic field strength data of the mutual inductor at different current and voltage change rates, extracts the hysteresis effect characteristics in combination with the hysteresis loop diagram, establishes an error model, and evaluates the hysteresis of the mutual inductor; and uses digital signal processing technology to perform spectrum analysis on the output signal of the mutual inductor to identify and predict harmonic distortion. Through a comprehensive analysis of the hysteresis of the hysteresis effect and the harmonic distortion anomaly index, the signal accuracy of the compensation algorithm is evaluated using a machine learning model, thereby achieving effective correction of high-frequency nonlinear errors. According to the accuracy of the compensation algorithm, the signal is divided into a high-accuracy signal and a low-accuracy signal. The high-accuracy signal is directly output, and the low-accuracy signal is output after error correction.

[0037] 2. The present invention effectively solves the problem that conventional compensation algorithms in the prior art cannot cope with the nonlinear errors of transformers, especially in high-frequency and extreme electromagnetic environments. It can significantly improve the accuracy and stability of transformer output signals and reduce the risk of system instability caused by error accumulation. Through the adaptive optimization of the machine learning model, the adaptability of the compensation algorithm is further improved, ensuring the long-term reliable operation of the system under complex working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0039] Figure 1 The present invention is a flow chart of the method.

[0040] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] Example 1, please refer to Figure 1 and Figure 2 As shown, the efficient and high-precision compensation method for an intelligent mutual inductor described in this embodiment includes the following steps:

[0043] S1: Under different current and voltage change rates, the magnetic flux density and magnetic field strength data of the transformer are collected to generate a hysteresis loop diagram;

[0044] S2: According to the hysteresis loop diagram, the hysteresis loop width characteristics and the residual magnetism and coercive force characteristics in the hysteresis loop are extracted respectively, and the hysteresis effect error model of the mutual inductor is established. The hysteresis of the hysteresis effect is determined according to the output results of the model;

[0045] S3: Use digital signal processing to perform spectrum analysis on the output signal of the transformer, identify the harmonic distortion of different frequency components in the signal, and predict the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment;

[0046] S4: Comprehensively analyze the hysteresis of the hysteresis effect and the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment to evaluate the signal accuracy of the compensation algorithm when dealing with the nonlinear error of the transformer;

[0047] S5: The signal accuracy of the compensation algorithm in response to the nonlinear error of the transformer is divided into different levels, that is, high-accuracy signals and low-accuracy signals. The high-accuracy signal is directly used as the final output signal of the transformer, and the low-accuracy signal is output after error correction.

[0048] Among them, in S1, under different current and voltage change rates, the magnetic flux density and magnetic field strength data of the transformer are collected to generate a hysteresis loop diagram, specifically:

[0049] Select the transformer to be tested, and its operating conditions should cover different current and voltage change rates to ensure that the experimental results have wide applicability. High-precision power supply for providing variable voltage and current. The power supply should have the ability to operate at different frequencies and voltages to support voltage and current changes at different rates. Sensors include: Current sensor: used to monitor the input current changes of the transformer in real time. Magnetic flux density sensor (Hall sensor): placed in or around the core of the transformer to measure the magnetic flux density (B). The Hall sensor can sense the changes in the magnetic field of the transformer and obtain flux information. Magnetic field strength sensor (current clamp, BH curve tester): used to measure the magnetic field strength (H) of the transformer. The current clamp is usually connected to the excitation power supply to monitor the magnetic field strength in the core of the transformer. Collect sensor data in real time and transfer the data to a computer for analysis. The data acquisition system should support high sampling rates to ensure accurate capture of magnetic field characteristics under rapidly changing voltage and current conditions.

[0050] According to the actual application scenario, set the operating current and voltage range of the transformer. Common voltage change rates can range from 0.1V / s to several hundred V / s, and current change rates can range from low frequency (such as 50Hz) to high frequency (such as several thousand Hz). The current and voltage change rates should cover from normal working conditions to extreme conditions to fully evaluate the hysteresis effect of the transformer. Before the experiment begins, calibrate all sensors to ensure the accuracy of equipment such as Hall sensors and current clamps, eliminate noise interference in the system, and ensure the accuracy of the measurement.

[0051] Apply current and voltage of different amplitudes gradually, and control the voltage change rate and current change rate. For example, start from a low voltage, increase at a set voltage rate, gradually apply to a high voltage, and record the corresponding changes in magnetic field strength and magnetic flux density. Start from a lower voltage and current change rate (such as low frequency and low voltage), apply a smaller magnetic field strength, and record the relationship between magnetic field strength and magnetic flux density. At this stage, the transformer core generally does not show obvious saturation, and the measured data is relatively linear. Gradually increase the amplitude and change rate of voltage and current until the transformer core gradually approaches or enters the magnetic saturation state. At this point, the relationship between magnetic field strength and magnetic flux density becomes nonlinear, and the hysteresis effect will be more obvious. In order to generate a complete hysteresis loop, the application of voltage and current should be changed in positive and negative cycles. For example, apply a positive voltage and gradually reduce it, then apply a negative voltage, repeat this process, and collect data to generate a closed hysteresis loop. At each current and voltage change rate, collect magnetic field strength (H) and magnetic flux density (B) data and record them at a high sampling rate. When the voltage and current are switched positive and negative, accurate recording of each sampling point is ensured to form a complete magnetization loop.

[0052] The magnetic field strength and flux density data collected during the experiment may contain noise and need to be preprocessed. Common preprocessing steps include: Filtering: Use digital filtering technology (such as low-pass filter or Kalman filter) to remove high-frequency noise to ensure smoother data. Correct zero drift or offset in the data to ensure that the starting point of the measurement is consistent with the reference value. Since the collected data points may be sparse under certain voltage change rates, interpolation algorithms (such as linear interpolation and spline interpolation) can be used to supplement the intermediate points to generate a smoother curve.

[0053] The processed magnetic field intensity (H) and magnetic flux density (B) data are visualized to generate a hysteresis loop diagram. The loop should be a closed loop curve, which represents the relationship between the magnetic field intensity and the magnetic flux density during the magnetization and demagnetization process of the transformer. Magnetization process: When the current or voltage gradually increases, the magnetic flux density increases with the magnetic field intensity, and the positive half of the hysteresis loop is formed. When the current or voltage gradually decreases, the relationship between the magnetic flux density and the magnetic field intensity lags, forming the negative half of the hysteresis loop.

[0054] The characteristic parameters of the hysteresis effect of the transformer are extracted from the generated hysteresis loop diagram, including: Hysteresis loop width: measures the width of the hysteresis loop, reflecting the energy loss in the iron core. The larger the width, the more serious the hysteresis effect. Remanence (Br): When the magnetic field strength (H) is reduced to zero, the remaining value of the magnetic flux density in the loop, indicating the hysteresis characteristics of the hysteresis effect. Coercive force (Hc): The reverse magnetic field strength required to return the magnetic flux density to zero, reflecting the closure of the hysteresis loop. The magnetic saturation of the transformer core is evaluated by observing the shape change of the loop. When the core is saturated, the top of the loop will tend to be flat, indicating that the magnetic flux density no longer increases linearly with the magnetic field strength. This saturation phenomenon will affect the accuracy of the transformer.

[0055] Repeat the above steps to generate multiple sets of hysteresis loops under different frequencies (such as 50Hz, 1kHz, etc.) and different loads (such as light load, heavy load). Under different frequency and load conditions, the hysteresis effect may show different characteristics, and the shape and width of the loop will change. Compare the loop diagrams under different conditions to analyze the differences in the hysteresis effect of the mutual inductor under various working conditions. The higher the frequency and the larger the load, the more obvious the hysteresis effect will be.

[0056] The collected raw data and processed hysteresis loop are saved for subsequent compensation algorithm input and system optimization. Based on the experimental results, an analysis report is generated to describe in detail the hysteresis effect characteristics of the transformer at different current and voltage change rates, as well as the magnetic saturation situation. These data can be used for subsequent compensation algorithm design and nonlinear error correction.

[0057] In this application, the magnetic flux density and magnetic field strength data of the transformer are collected at different current and voltage change rates, which can generate accurate hysteresis loop diagrams and extract key parameters from them. These data can not only help evaluate the hysteresis effect and nonlinear characteristics of the transformer, but also provide basic data for the subsequent compensation algorithm design, ensuring that the system can effectively cope with complex working conditions.

[0058] S2: According to the hysteresis loop diagram, the hysteresis loop width characteristics and the residual magnetism and coercive force characteristics in the hysteresis loop are extracted respectively, and the hysteresis effect error model of the mutual inductor is established. The hysteresis of the hysteresis effect is determined according to the output results of the model.

[0059] The width of the hysteresis loop indicates the degree of hysteresis between the magnetic field intensity H and the magnetic flux density B during the magnetization and demagnetization process of the transformer, and is an important feature of the hysteresis effect. The width of the hysteresis loop can be calculated by measuring the difference in magnetic field intensity when the magnetic flux density is zero.

[0060] In the hysteresis loop diagram, find the point where the magnetic flux density B=0. When B=0, record the positive and negative magnetic field strengths, recorded as Hpositive and Hnegative respectively, and calculate the expression: Wh=Hpositive-Hnegative; where: Wh is the width of the hysteresis loop, reflecting the hysteresis in the magnetization and demagnetization process, Hpositive is the positive magnetic field strength when the magnetic flux density is zero, and Hnegative is the negative magnetic field strength when the magnetic flux density is zero.

[0061] Remanence is the value of magnetic flux density B when magnetic field strength H = 0. It indicates the magnetic flux retained in the core of the transformer after demagnetization. In the hysteresis loop diagram, find the point where magnetic field strength H = 0, record the magnetic flux density value at this time, record it as Br, and calculate the expression: Br = B (H = 0); where Br is remanence, which indicates the magnetic flux density in the magnetization loop when the magnetic field strength is zero.

[0062] Coercive force is the reverse magnetic field strength required to make the magnetic flux density B = 0. It indicates the magnetic field strength required by the transformer core to eliminate residual magnetism, and is an important feature of the hysteresis effect. In the hysteresis loop diagram, find the point where the magnetic flux density B = 0, record the corresponding reverse magnetic field strength, recorded as Hc, and calculate the expression: Hc = |H(B = 0)|; where Hc is the coercive force, which indicates the reverse magnetic field strength required to return the magnetic flux density to zero.

[0063] The hysteresis effect error model reflects the nonlinear error of the hysteresis effect in the operation of the transformer. The hysteresis effect error model is established through the hysteresis loop width Wh, remanence Br, and coercive force Hc. The hysteresis effect error is described as the hysteresis relationship between the input magnetic field intensity and the actual output magnetic flux density. Under a given magnetic field intensity H, the error ε caused by the hysteresis effect mag (H) The expression is: mag (H) = α·Wh + β·Br + γ·Hc; where α, β, and γ are unknown coefficients, which are determined by fitting experimental data and reflect the contribution of different parameters to the hysteresis error.

[0064] Through multiple experiments to collect data, calculate Wh, Br, Hc under different currents and voltage change rates, and then combine the actual measured transformer output signal error to determine the coefficients α, β, and γ using data fitting methods (such as the least squares method). The fitting steps are as follows: Experimental data collection: Collect Wh, Br, Hc under multiple different working conditions and the actual measured transformer output error. Use multiple regression or least squares method to fit the error formula and solve α, β, and γ. Repeatedly adjust the fitting process to ensure that the model has high accuracy in predicting errors under various conditions.

[0065] The error model can be used to determine the hysteresis of the transformer, that is, the influence of the hysteresis effect of the transformer on the output signal under different working conditions.

[0066] Hysteresis effect τ mag Through the maximum error estimation of the error model, the influence of hysteresis on the output under different current and voltage changes is calculated, and the expression is: τ mag =maxLε mag (H)M; compare the calculated hysteresis effect with the preset hysteresis effect hysteresis reference threshold value. If the hysteresis effect is greater than or equal to the hysteresis reference threshold value, it means that the hysteresis effect is serious, and a hysteresis effect hysteresis signal is generated. If the hysteresis effect is less than the hysteresis reference threshold value, it means that the hysteresis effect is not serious, and no hysteresis effect hysteresis signal is generated.

[0067] S3: Use digital signal processing to perform spectrum analysis on the output signal of the transformer, identify the harmonic distortion of different frequency components in the signal, and predict the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment.

[0068] The output signal of the transformer is collected in real time. In order to accurately capture the harmonic components of the signal, the sampling frequency fs must satisfy the Nyquist sampling theorem, that is, the sampling frequency is at least twice the highest frequency component of the signal. Usually, the selected sampling frequency should be greater than 10 times the fundamental frequency. For example, if the fundamental frequency of the power system is 50Hz, the sampling frequency should be at least 500Hz, or even higher. In order to obtain a high-resolution spectrum, it is important to collect sufficiently long signal samples. The signal length T is related to the spectrum resolution Δf, and Therefore, the longer the signal acquisition time T is, the higher the spectrum resolution is. When acquiring signals, pay attention to processing environmental noise and interference signals. You can use analog or digital filters (such as low-pass filters) to pre-process the signal to reduce high-frequency noise and interference in the sampled signal.

[0069] Before spectrum analysis of the collected transformer output signal, the DC offset must be removed first because the signal may contain a DC component to ensure that the spectrum analysis is focused on the AC part. The DC component can be eliminated by removing the mean value of the signal. The expression is: Where x(t) is the original signal and N is the number of signal samples. To reduce the spectrum leakage effect, a window function can be applied to the signal. Commonly used window functions include Hanning window, Hamming window and Heyman window. The choice of window function helps to enhance the resolution of the spectrum within a limited sampling length. The expression is: x(t) windowed=x(t)·w(t); where w(t) is the window function. The preprocessed time domain signal x(t) is subjected to a fast Fourier transform operation to obtain a frequency domain signal X(f), which is expressed as: X(k) is the kth discrete frequency component in the frequency domain, N is the total number of sampling points, k represents different frequency components, x(n) is the nth sampling point in the time domain signal, is the kernel function of Fourier transform. Through fast Fourier transform, the amplitude and phase information of the signal at different frequencies are obtained. Harmonic components usually appear at integer multiples of the fundamental frequency f0, such as 2f0 (second harmonic), 3f0 (third harmonic), etc.

[0070] After obtaining the frequency domain data through the fast Fourier transform operation, a spectrum is drawn with the horizontal axis being the frequency (Hz) and the vertical axis being the signal amplitude (or power density). In the spectrum, the harmonic signal will appear as a peak at an integer multiple of the fundamental frequency f0. Usually, the fundamental frequency of the power system is 50Hz or 60Hz, so the second, third, and fifth harmonics appear at 100Hz, 150Hz, 250Hz, etc.

[0071] Through spectrum analysis, the harmonic distortion in the signal is identified and quantified. Total harmonic distortion is an important indicator to measure harmonic distortion. It represents the ratio of the sum of the squares of all harmonic components in the signal to the square of the fundamental component. The expression is: Where THD is the total harmonic distortion, V1 is the amplitude of the fundamental frequency (50Hz or 60Hz), and V8, V9, ... are the amplitudes of each harmonic (such as the second harmonic, the third harmonic, etc.).

[0072] The higher the THD, the more harmonic components there are in the signal and the more serious the distortion. To identify the frequency components and amplitudes of each specific harmonic, we usually pay attention to the amplitudes of low-order harmonics (such as second and third harmonics). These harmonics often have a greater impact on the measurement accuracy of the transformer. The amplitude of each harmonic V Q It can be obtained through spectrum analysis.

[0073] The harmonic distortion degree of the output signal of the transformer in a high-frequency environment is predicted to generate a harmonic distortion anomaly index. The method for obtaining the harmonic distortion anomaly index is as follows:

[0074] Select wavelet basis functions, such as Daubechies (db), Symlets (sym), Coiflets (coif), etc., perform multi-scale wavelet decomposition on the harmonic distortion data, and obtain detail coefficients and approximate coefficients at different scales. The specific calculation expression is: Where, X(t) is the transformer output signal, φ j％,- (t) and ψ j,-(t) are the scaling function and wavelet function, respectively, A j％,- is the approximation coefficient, D j,- is the detail coefficient;

[0075] The signal is reconstructed using detail coefficients and approximate coefficients to obtain reconstructed signals at different scales. The specific calculation expression is: Among them, AX j％,- and DX j,- are the reconstructed approximate coefficient and detail coefficient respectively, and BX(t) is the reconstructed signal; the harmonic distortion anomaly index is calculated according to the reconstructed signal, and the variation of the output signal of the transformer is obtained by calculating the energy change of the detail coefficient at different scales, and the energy of the detail coefficient at each scale is calculated. The specific calculation expression is: E j =∑ - AD j,- A 8 ; Among them, E j is the energy of the jth layer. The harmonic distortion anomaly index is calculated according to the energy changes at different scales. The specific calculation expression is: Where QV is the harmonic distortion anomaly index.

[0076] The larger the harmonic distortion anomaly index is, the higher the proportion of high-frequency harmonic components in the transformer output signal to the total signal energy. This means that the more serious the harmonic distortion of the signal is, especially in a high-frequency environment, the harmonic distortion may increase significantly, affecting the signal quality and measurement accuracy. A high harmonic distortion anomaly index usually means that there is a large nonlinear load in the system or that the electrical equipment is unstable, which may lead to equipment misoperation, measurement errors, and even power system instability.

[0077] On the contrary, the smaller the harmonic distortion abnormality index is, the less high-frequency harmonic components of the transformer output signal are, and the harmonic distortion of the signal is lower. In this case, the output signal of the transformer is closer to the ideal fundamental signal, the measurement accuracy is higher, and the system operation is more stable. This usually means that the influence of high-frequency harmonics can be ignored, the signal quality is better, and the power system is in a normal and stable working state.

[0078] S4: Comprehensively analyze the hysteresis of the hysteresis effect and the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment to evaluate the signal accuracy of the compensation algorithm when dealing with the nonlinear error of the transformer.

[0079] The hysteresis effect and the harmonic distortion anomaly index are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model uses each group of first eigenvectors to predict the signal accuracy value label of the compensation algorithm when dealing with the transformer nonlinear error as the prediction target, and takes minimizing the sum of the prediction errors of the signal accuracy value labels of all compensation algorithms when dealing with the transformer nonlinear error as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The signal accuracy value of the compensation algorithm when dealing with the transformer nonlinear error is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0080] The method for obtaining the signal accuracy value when the compensation algorithm copes with the nonlinear error of the mutual inductor is as follows: from the first eigenvector training data of the trained machine learning model, the corresponding function expression is obtained: CR = FUτ mag , QVV; where F is the output function of the model, τ mag is the hysteresis of the hysteresis effect, QV is the harmonic distortion anomaly index, and CR is the signal accuracy value when the compensation algorithm deals with the nonlinear error of the transformer.

[0081] S5: The signal accuracy of the compensation algorithm in response to the nonlinear error of the transformer is divided into different levels, that is, high-accuracy signals and low-accuracy signals. The high-accuracy signal is directly used as the final output signal of the transformer, and the low-accuracy signal is output after error correction.

[0082] The obtained signal accuracy value of the compensation algorithm in response to the nonlinear error of the transformer is compared with the pre-set reference threshold of the signal accuracy value. If the signal accuracy value of the compensation algorithm in response to the nonlinear error of the transformer is greater than or equal to the reference threshold of the signal accuracy value, it means that the signal accuracy of the compensation algorithm in response to the nonlinear error of the transformer is high. At this time, no warning signal is generated, and the signal of the compensation algorithm in response to the nonlinear error of the transformer is classified as a high-accuracy signal, which is directly used as the final output signal of the transformer; if the signal accuracy value of the compensation algorithm in response to the nonlinear error of the transformer is less than the reference threshold of the signal accuracy value, it means that the signal accuracy of the compensation algorithm in response to the nonlinear error of the transformer is low. At this time, a warning signal is generated, and the signal of the compensation algorithm in response to the nonlinear error of the transformer is classified as a low-accuracy signal, which is output after error correction.

[0083] When the output signal of the transformer is evaluated as a high-accuracy signal, it indicates that its error is within the allowable range and meets the predetermined accuracy standard. In this case, there is no need to further process or correct the signal, and it can be directly used as the final output signal of the transformer and transmitted to the control center or monitoring equipment of the power system. This processing method can ensure the rapid transmission of signals and reduce unnecessary delays, while maintaining efficient operation and accurate measurement of the system, especially in scenarios that require real-time response, such as power dispatching or automation control systems.

[0084] When the output signal of the transformer is evaluated as low accuracy, the error of the signal exceeds the set tolerance range and must be corrected before output to prevent the erroneous signal from affecting the safety and stability of the power system. The error correction process includes the following steps:

[0085] Signal analysis: First, analyze the low-accuracy signal in detail to identify the source of the error. Common error types include: Harmonic distortion: There are many high-order harmonic components, which cause serious signal distortion.

[0086] Hysteresis effect: The transformer core saturates under high load or high voltage conditions, resulting in increased nonlinearity errors.

[0087] Noise interference: Environmental noise or electromagnetic interference introduces high-frequency noise components.

[0088] Spectrum analysis and model detection: Use tools such as fast Fourier transform (FFT) or wavelet transform to perform spectrum analysis on the signal, determine the frequency characteristics of harmonic or noise interference, and use the hysteresis effect model to make error judgments.

[0089] If there is significant harmonic distortion in the signal, the harmonics in a specific frequency band can be suppressed by applying a digital filter (such as a band-stop filter or a low-pass filter). Digital filters can effectively remove high-frequency noise and high-order harmonics and correct the distorted part of the signal. Adaptive filtering: For situations where the harmonic frequency is not fixed, an adaptive filter can be used to dynamically adjust the frequency response of the filter and correct the high-frequency noise and distortion components of the signal in real time.

[0090] For nonlinear errors (such as nonlinear response caused by hysteresis effect), nonlinear compensation algorithms (such as fuzzy logic control, neural network, etc.) can be used to dynamically compensate signal output by analyzing the error model. Neural networks can automatically adjust compensation parameters by training historical data, and fuzzy logic can handle fuzzy error characteristics.

[0091] If the signal contains random noise or multiple interferences, the Kalman filter can correct the signal by making the best estimate of the noise. It is suitable for continuously correcting low-precision signals in dynamic systems to make the output signal smoother and more accurate.

[0092] The corrected signal is monitored in real time through the feedback loop to ensure that the error-corrected signal meets the requirements. If the error correction does not achieve the expected effect, the feedback system will trigger readjustment until the signal accuracy reaches the standard. By comparing with a high-precision reference signal or model, it is verified whether the corrected signal is within the predetermined error range, ensuring that the compensated signal can be safely output.

[0093] After the signal is error corrected, the error is reduced to an acceptable range, and the corrected signal is then used as the final output of the transformer and transmitted to the control system or monitoring platform to ensure safe operation and stable control of the system.

[0094] In this application, through this double-layer processing mechanism, high-accuracy signals can be directly used as the final output of the transformer, ensuring efficient signal transmission and rapid response of the system. For low-accuracy signals, after various error correction methods such as harmonic filtering, nonlinear compensation, and adaptive correction, the accuracy of the signal before output is improved. Ultimately, this solution can significantly improve the measurement accuracy and reliability of the transformer under complex working conditions and ensure the stable operation of the power system.

[0095] In this embodiment, under different current and voltage change rates, magnetic flux density and magnetic field strength data are collected to generate a hysteresis loop diagram; the hysteresis loop width, residual magnetism and coercive force characteristics are extracted according to the hysteresis loop, and a hysteresis effect error model is established to determine the hysteresis; digital signal processing is used to perform spectrum analysis on the output signal to identify and predict the degree of harmonic distortion; a comprehensive analysis is performed on the hysteresis effect and the degree of harmonic distortion to evaluate the signal accuracy of the compensation algorithm; and graded processing is performed according to the signal accuracy: the high-accuracy signal is directly used as the final output, and the low-accuracy signal is output after error correction. This method ensures high-precision measurement and signal output of the mutual inductor under complex working conditions.

[0096] Embodiment 2, a high-efficiency and high-precision compensation system for an intelligent mutual inductor described in this embodiment includes a data acquisition module, a hysteresis effect modeling module, a spectrum analysis and harmonic detection module, a comprehensive error analysis and evaluation module, and a signal classification and output module;

[0097] Data acquisition module: collects the flux density and magnetic field strength data of the transformer at different current and voltage change rates to generate a hysteresis loop diagram;

[0098] Hysteresis effect modeling module: According to the hysteresis loop diagram, the hysteresis loop width characteristics, residual magnetism and coercive force characteristics in the hysteresis loop are extracted respectively, and the hysteresis effect error model of the mutual inductor is established. The hysteresis of the hysteresis effect is determined according to the output results of the model;

[0099] Spectrum analysis and harmonic detection module: Use digital signal processing to perform spectrum analysis on the output signal of the transformer, identify the harmonic distortion of different frequency components in the signal, and predict the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment;

[0100] Comprehensive error analysis and evaluation module: Comprehensively analyze the hysteresis of the hysteresis effect and the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment, and evaluate the signal accuracy of the compensation algorithm when dealing with the nonlinear error of the transformer;

[0101] Signal classification and output module: The signal accuracy when the compensation algorithm responds to the nonlinear error of the transformer is divided into different levels, that is, high-accuracy signals and low-accuracy signals. The high-accuracy signal is directly used as the final output signal of the transformer, and the low-accuracy signal is output after error correction.

[0102] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0103] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0104] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0105] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. An efficient and high-precision compensation method for an intelligent mutual inductor, characterized in that: The steps include: S1: Under different current and voltage change rates, the magnetic flux density and magnetic field strength data of the transformer are collected to generate a hysteresis loop diagram; S2: According to the hysteresis loop diagram, the hysteresis loop width characteristics and the residual magnetism and coercive force characteristics in the hysteresis loop are extracted respectively, and the hysteresis effect error model of the mutual inductor is established. The hysteresis of the hysteresis effect is determined according to the output results of the model; S3: Use digital signal processing to perform spectrum analysis on the output signal of the transformer, identify the harmonic distortion of different frequency components in the signal, and predict the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment; S4: Comprehensively analyze the hysteresis of the hysteresis effect and the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment to evaluate the signal accuracy of the compensation algorithm when dealing with the nonlinear error of the transformer; S5: The signal accuracy of the compensation algorithm in response to the nonlinear error of the transformer is divided into different levels, that is, high-accuracy signals and low-accuracy signals. The high-accuracy signal is directly used as the final output signal of the transformer, and the low-accuracy signal is output after error correction.

2. The high-efficiency and high-precision compensation method for an intelligent mutual inductor according to claim 1 is characterized in that: In S2, according to the hysteresis loop diagram, the hysteresis loop width characteristics and the residual magnetism and coercive force characteristics in the hysteresis loop are extracted respectively, and the hysteresis effect error model of the mutual inductor is established. The hysteresis of the hysteresis effect is determined according to the output results of the model, specifically: The hysteresis loop width indicates the degree of hysteresis between the magnetic field intensity H and the magnetic flux density B during the magnetization and demagnetization process of the transformer, and is calculated by measuring the difference in magnetic field intensity when the magnetic flux density is zero; In the hysteresis loop diagram, find the point where the magnetic flux density B = 0. When B = 0, record the positive and negative magnetic field strengths, record them as Hpositive and Hnegative respectively, and calculate the expression: Wh = Hpositive-Hnegative; where: Wh is the width of the hysteresis loop, reflecting the hysteresis in the magnetization and demagnetization process, Hpositive is the positive magnetic field strength when the magnetic flux density is zero, and Hnegative is the negative magnetic field strength when the magnetic flux density is zero; Remanence is the value of magnetic flux density B when magnetic field strength H=0, indicating the magnetic flux retained in the core of the transformer after demagnetization. In the hysteresis loop diagram, find the point where magnetic field strength H=0, record the magnetic flux density value at this time, recorded as Br, and calculate the expression: Br=B(H=0); where Br is remanence, indicating the magnetic flux density in the magnetization loop when the magnetic field strength is zero; Coercive force is the reverse magnetic field strength required to make the magnetic flux density B = 0, which indicates the magnetic field strength required by the transformer core to eliminate residual magnetism. In the hysteresis loop diagram, find the point where the magnetic flux density B = 0, record the corresponding reverse magnetic field strength, recorded as Hc, and calculate the expression: Hc = |H(B = 0)|; where Hc is the coercive force, which indicates the reverse magnetic field strength required to return the magnetic flux density to zero; The hysteresis effect error model reflects the nonlinear error of the hysteresis effect in the operation of the transformer. The hysteresis effect error model is established through the hysteresis loop width Wh, remanence Br, and coercive force Hc. The hysteresis effect error is described as the hysteresis relationship between the input magnetic field intensity and the actual output magnetic flux density. Under a given magnetic field intensity H, the error ε caused by the hysteresis effect is mag (H) The expression is: mag (H) = α·Wh+β·Br+γ·Hc; wherein α, β, and γ are unknown coefficients.

3. The high-efficiency and high-precision compensation method for an intelligent mutual inductor according to claim 2 is characterized in that: In S3, digital signal processing is used to perform spectrum analysis on the output signal of the transformer, identify the harmonic distortion of different frequency components in the signal, and predict the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment, specifically: The output signal of the transformer is collected in real time. Before spectrum analysis, the DC component is eliminated by removing the mean value of the signal. The expression is: Where x(t) is the original signal, N is the number of signal samples, and a window function is applied to the signal to enhance the resolution of the spectrum within a limited sampling length. The expression is: windowed =x(t)·w(t); where w(t) is a window function; the preprocessed time domain signal x(t) is subjected to a fast Fourier transform operation to obtain a frequency domain signal X(f), which is expressed as: X(k) is the kth discrete frequency component in the frequency domain, N is the total number of sampling points, k represents different frequency components, x(n) is the nth sampling point in the time domain signal, is the kernel function of Fourier transform. After obtaining the frequency domain data through fast Fourier transform operation, a spectrum diagram is drawn, with the horizontal axis being the frequency and the vertical axis being the signal amplitude. In the spectrum diagram, the harmonic signal will appear as a peak value at an integer multiple of the fundamental frequency f0; Through spectrum analysis, the harmonic distortion in the signal is identified and quantified, which is the ratio of the sum of the squares of all harmonic components in the signal to the square of the fundamental component. The expression is: Where THD is the total harmonic distortion, V1 is the amplitude of the fundamental frequency, and V8, V9, ... are the amplitudes of each harmonic.

4. The high-efficiency and high-precision compensation method for an intelligent mutual inductor according to claim 3 is characterized in that: The harmonic distortion degree of the output signal of the transformer in a high-frequency environment is predicted to generate a harmonic distortion anomaly index. The method for obtaining the harmonic distortion anomaly index is as follows: Select wavelet basis function, perform multi-scale wavelet decomposition on harmonic distortion data, obtain detail coefficients and approximate coefficients at different scales, use detail coefficients and approximate coefficients to reconstruct the signal, and obtain reconstructed signals at different scales. The specific calculation expression is: BX(t) = Among them, AX j％,- and DX j,- are the reconstructed approximate coefficient and detail coefficient respectively, and BX(t) is the reconstructed signal; the harmonic distortion anomaly index is calculated according to the reconstructed signal, and the variation of the output signal of the transformer is obtained by calculating the energy change of the detail coefficient at different scales, and the energy of the detail coefficient at each scale is calculated. The specific calculation expression is: E j =∑ - AD j,- A 8 ; Among them, E j is the energy of the jth layer. The harmonic distortion anomaly index is calculated according to the energy changes at different scales. The specific calculation expression is: Where QV is the harmonic distortion anomaly index.

5. The high-efficiency and high-precision compensation method for an intelligent mutual inductor according to claim 4 is characterized in that: In S4, the hysteresis of the hysteresis effect and the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment are comprehensively analyzed, specifically: The hysteresis effect and the harmonic distortion anomaly index are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model uses each group of first eigenvectors to predict the signal accuracy value label of the compensation algorithm when dealing with the transformer nonlinear error as the prediction target, and takes minimizing the sum of the prediction errors of the signal accuracy value labels of all compensation algorithms when dealing with the transformer nonlinear error as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The signal accuracy value of the compensation algorithm when dealing with the transformer nonlinear error is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

6. The high-efficiency and high-precision compensation method for an intelligent mutual inductor according to claim 1, characterized in that: In S5, the signal accuracy of the compensation algorithm in dealing with the nonlinear error of the mutual inductor is divided into different levels, which are divided into high-accuracy signals and low-accuracy signals, specifically: The obtained signal accuracy value of the compensation algorithm in response to the nonlinear error of the mutual inductor is compared with the preset reference threshold of the signal accuracy value. If the signal accuracy value of the compensation algorithm in response to the nonlinear error of the mutual inductor is greater than or equal to the reference threshold of the signal accuracy value, it means that the signal accuracy of the compensation algorithm in response to the nonlinear error of the mutual inductor is high. At this time, no warning signal is generated, and the signal of the compensation algorithm in response to the nonlinear error of the mutual inductor is classified as a high-accuracy signal, which is directly used as the final output signal of the mutual inductor; If the signal accuracy value when the compensation algorithm responds to the nonlinear error of the transformer is less than the reference threshold value of the signal accuracy value, it means that the signal accuracy when the compensation algorithm responds to the nonlinear error of the transformer is low. At this time, a warning signal is generated, and the signal when the compensation algorithm responds to the nonlinear error of the transformer is classified as a low-accuracy signal, which is output after error correction.

7. An efficient and high-precision compensation system for an intelligent mutual inductor, used to implement an efficient and high-precision compensation method for an intelligent mutual inductor according to any one of claims 1 to 6, characterized in that: It includes data acquisition module, hysteresis effect modeling module, spectrum analysis and harmonic detection module, comprehensive error analysis and evaluation module, and signal classification and output module; Data acquisition module: collects the flux density and magnetic field strength data of the transformer at different current and voltage change rates to generate a hysteresis loop diagram; Hysteresis effect modeling module: According to the hysteresis loop diagram, the hysteresis loop width characteristics, residual magnetism and coercive force characteristics in the hysteresis loop are extracted respectively, and the hysteresis effect error model of the mutual inductor is established. The hysteresis of the hysteresis effect is determined according to the output results of the model; Spectrum analysis and harmonic detection module: Use digital signal processing to perform spectrum analysis on the output signal of the transformer, identify the harmonic distortion of different frequency components in the signal, and predict the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment; Comprehensive error analysis and evaluation module: Comprehensively analyze the hysteresis of the hysteresis effect and the degree of harmonic distortion of the output signal of the transformer in a high-frequency environment, and evaluate the signal accuracy of the compensation algorithm when dealing with the nonlinear error of the transformer; Signal classification and output module: The signal accuracy when the compensation algorithm responds to the nonlinear error of the transformer is divided into different levels, that is, high-accuracy signals and low-accuracy signals. The high-accuracy signal is directly used as the final output signal of the transformer, and the low-accuracy signal is output after error correction.

Images