A data processing method and system based on current transformer ratio error

Abstract

This invention provides a data processing method and system based on the ratio difference of a current transformer, belonging to the field of current transformer technology. The method includes: synchronously acquiring the primary and secondary current signals of the transformer, obtaining a current waveform dataset after aperiodic component subtraction and wavelet denoising; calibrating the dual electromagnetic circuits of the magnetic valve-type transformer separately, completing the error data through bicubic interpolation and fitting an error function model; extracting the amplitude and phase angles of the fundamental and harmonic waves through ChirpZ transform, and obtaining the frequency domain error by combining FFT transform with the error model; calculating the initial ratio difference and compensating for harmonic flux linkage, inversely transforming the frequency domain error to the time domain, and weighted synthesizing to obtain the final ratio difference dataset, thereby improving the ratio difference measurement accuracy and adapting to the structural characteristics of the dual electromagnetic circuits.

Description

Technical Field

[0001] This invention belongs to the field of current transformer technology, and more specifically, relates to a data processing method and system based on the ratio difference of current transformers. Background Technology

[0002] Current transformers are devices used for fault protection and condition monitoring in power systems, and the accuracy of their differential measurement is crucial to the economy and safety of power grid operation. With the advancement of smart grid construction, magnetic valve-type current transformers have been widely used due to their advantages of wide bandwidth and resistance to DC saturation. These transformers have a special structure that includes a magnetic field sensor and a dual electromagnetic circuit for the secondary winding.

[0003] Traditional current transformer differential data processing methods rely heavily on standard transformer comparison measurements, which are cumbersome and subject to significant environmental constraints. They also fail to specifically remove the exponentially decaying non-periodic components in the signal, causing interference to affect the accuracy of subsequent feature extraction. Furthermore, they cannot adapt to the dual-loop structure characteristics of magnetic valve-type current transformers and do not perform separate calibration measurements on the magnetic field sensor circuit and secondary winding circuit. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a data processing method and system based on the ratio difference of current transformers. This addresses the issue that traditional data processing methods often rely on comparative measurements with standard transformers, failing to specifically deduct the exponentially decaying aperiodic components in the signal, thus causing interference in the signal to affect the accuracy of subsequent feature extraction.

[0005] The purpose and effectiveness of the data processing method and system based on the ratio difference of current transformers of the present invention are achieved by the following specific technical means:

[0006] A data processing method based on the ratio difference of current transformers includes the following steps:

[0007] S1: Connect the data acquisition card to the primary and secondary current detection terminals of the current transformer to obtain the current waveform dataset;

[0008] S2: Perform dual electromagnetic loop calibration measurement based on current waveform dataset to obtain measurement error datasets for the two electromagnetic loops. Perform bicubic interpolation and fitting based on the measurement error datasets for the two electromagnetic loops to obtain error function model dataset.

[0009] S3: Based on the current waveform dataset, refine the fundamental and harmonic amplitude and phase angle datasets of the first and second sides, perform FFT transformation on the current waveform dataset and fit it with the error function model dataset to obtain signal error datasets of different frequencies in the frequency domain of the two loops.

[0010] S4: Calculate the ratio difference based on the amplitude and phase angle datasets of the primary and secondary fundamental waves and harmonics respectively, and input the preset harmonic flux compensation model to obtain the ratio difference dataset and the ratio difference compensation vector dataset.

[0011] S5: Based on the ratio difference dataset and the ratio difference compensation vector dataset, perform vector summation, convert the signal error datasets of different frequencies in the frequency domain of the two circuits back to the time domain, and perform weighted synthesis with the vector summation result to obtain the current transformer ratio difference dataset.

[0012] According to a preferred embodiment, the step of connecting the acquisition card to the primary and secondary current detection terminals of the current transformer to obtain the current waveform dataset includes:

[0013] Deploy a synchronous data acquisition card, connect the acquisition card to the primary and secondary current detection terminals of the current transformer, set the sampling frequency, synchronously acquire the current signal of the transformer under transient or steady-state operating conditions, continuously acquire several power frequency cycles, and obtain the original current waveform dataset of the primary and secondary sides.

[0014] A current mathematical model containing sinusoidal components and exponentially decaying aperiodic components is established based on the original current waveform dataset of the primary and secondary sides. The model parameters are solved iteratively using the nonlinear least squares method to obtain the optimal fitting result of the aperiodic component. This component is then subtracted from the original data to obtain the corrected current waveform dataset. The corrected current waveform dataset is then decomposed to obtain the current waveform dataset.

[0015] According to a preferred embodiment, the step of decomposing the current waveform dataset based on the corrected current waveform dataset to obtain the current waveform dataset includes:

[0016] Based on the corrected current waveform dataset, the data is decomposed into 5 levels using the sym8 wavelet basis function to obtain the approximation coefficients and detail coefficients of each level. Based on the Stan unbiased risk estimation criterion, the optimal threshold is adaptively calculated for the detail coefficients of levels 1 to 5. The detail coefficients are processed by a soft threshold function, and the processed detail coefficients are reconstructed with the approximation coefficients of level 5 to obtain the current waveform dataset.

[0017] According to a preferred embodiment, the step of performing dual electromagnetic loop calibration measurement based on current waveform dataset to obtain measurement error datasets for the two electromagnetic loops, and then performing bicubic interpolation and fitting based on the measurement error datasets for the two electromagnetic loops to obtain an error function model dataset includes:

[0018] Based on the current waveform dataset, the magnetic field sensor section and the secondary winding section of the magnetic valve current transformer are selected respectively. The magnetic field sensor section is compared with the standard magnetic field sensor to measure the output signals of DC and AC signals of different frequencies. The secondary winding section is compared with the standard current transformer to measure the output signals under different current frequencies, and the measurement error datasets of the two electromagnetic circuits are obtained.

[0019] Based on the measurement error dataset of two electromagnetic loops, 16 measurement points around the target point are selected, the weight coefficient corresponding to the distance from each point to the target point is calculated, the error measurement value of each point is multiplied by the corresponding coefficient and then summed to supplement the error data in the frequency and amplitude range, and the interpolated error supplement dataset is obtained.

[0020] The error function model dataset is obtained by least squares fitting based on the interpolated error supplement dataset.

[0021] According to a preferred embodiment, the step of refining the current waveform dataset to obtain the amplitude and phase angle datasets of the primary and secondary fundamental waves and harmonics, performing an FFT transformation on the current waveform dataset and fitting it with the error function model dataset to obtain signal error datasets of different frequencies in the frequency domain of the two loops includes:

[0022] Based on the current waveform dataset, for the nominal frequencies of the fundamental wave and the 3rd, 5th and 7th harmonics, a narrow-band analysis frequency range centered on the nominal frequency is set. Chirp Z transform is performed within the narrow-band range to refine the spectrum. Secondary interpolation is performed on the peak points of the discrete spectrum and their adjacent points to obtain the amplitude and phase angle datasets of the fundamental wave and harmonics on the first and second sides.

[0023] Based on the current waveform dataset, the output signals of the magnetic field sensor section and the secondary winding section are converted into the frequency domain to obtain the amplitude-frequency datasets of the output signals of the two electromagnetic circuits. Based on the amplitude-frequency datasets of the output signals of the two electromagnetic circuits and the error function model dataset, the amplitude-frequency data are substituted into the error function model to calculate the signal error at different frequencies, and obtain the signal error datasets of the two circuits at different frequencies in the frequency domain.

[0024] According to a preferred embodiment, the step of calculating the ratio difference based on the amplitude and phase angle datasets of the primary and secondary fundamental waves and harmonics, and inputting them into a preset harmonic flux linkage compensation model to obtain the ratio difference dataset and the ratio difference compensation vector dataset includes:

[0025] Based on the primary and secondary fundamental and harmonic amplitude and phase angle datasets, the primary and secondary fundamental current amplitudes obtained by ChirpZ transform are extracted. At the same time, the nominal ratio calibrated at the time of manufacture of the magnetic valve current transformer is retrieved. The secondary fundamental current amplitude is multiplied by the nominal ratio to obtain the equivalent current amplitude of the secondary current converted to the primary side. The current amplitude deviation is obtained by subtracting the primary fundamental current amplitude from the equivalent current amplitude. The relative deviation ratio of the transformer ratio is obtained by dividing the current amplitude deviation by the primary fundamental current amplitude. Thus, the ratio difference dataset is obtained.

[0026] Based on the amplitude and phase angle datasets of the first and second fundamental waves and harmonics, the amplitudes and phase angles of the second fundamental wave and the third, fifth and seventh harmonics are input into a preset harmonic flux compensation model. The proportional coefficients in the harmonic flux compensation model are calibrated by a multiple linear regression algorithm, and the proportional difference compensation vector dataset is calculated.

[0027] According to a preferred embodiment, the step of performing vector summation based on the ratio difference dataset and the ratio difference compensation vector dataset, converting the signal error datasets of different frequencies in the frequency domain of the two circuits back to the time domain, and weighting and synthesizing them with the vector summation result to obtain the current transformer ratio difference dataset includes:

[0028] Based on the ratio difference dataset and the ratio difference compensation vector dataset, the initial ratio difference and the ratio difference compensation vector are vector summed to eliminate the additional error introduced by the harmonic components and obtain the compensated initial ratio difference dataset.

[0029] Inverse FFT transformation is performed on the signal error datasets of the two loops at different frequencies in the frequency domain to convert the frequency domain error data back to the time domain and obtain the time domain signal error datasets of the two electromagnetic loops.

[0030] Error synthesis is performed based on the initial ratio difference dataset after compensation and the time-domain signal error dataset of the two electromagnetic circuits. According to the proportion of the output signals of the magnetic field sensor section and the secondary winding section in the final measurement signal, the two types of error data are weighted and synthesized to obtain the current transformer ratio difference dataset.

[0031] A data processing system based on the ratio difference of current transformers includes:

[0032] The current waveform data acquisition module includes acquisition cards connected to the primary and secondary current detection terminals of the current transformer, used to acquire current waveform datasets.

[0033] The dual electromagnetic loop calibration and fitting module is connected to the current waveform data acquisition module. It is used to perform dual electromagnetic loop calibration measurement and fitting based on the current waveform dataset to obtain the error function model dataset.

[0034] The frequency domain error analysis module is connected to the current waveform data acquisition module and the dual electromagnetic loop calibration and fitting module. It is used to refine the current waveform dataset, and at the same time, it performs FFT transformation based on the current waveform dataset and fits it with the error function model dataset to obtain signal error datasets of different frequencies in the frequency domain of the two loops.

[0035] The compensation module, connected to the frequency domain error analysis module, includes a preset harmonic flux compensation model for acquiring the ratio difference compensation vector dataset;

[0036] The error synthesis module, connected to the compensation module and the frequency domain error analysis module, is used to perform vector summation on the ratio difference dataset and the ratio difference compensation vector dataset, convert the signal error datasets of different frequencies in the frequency domain of the two circuits back to the time domain, and perform weighted synthesis with the vector summation result to obtain the current transformer ratio difference dataset.

[0037] Compared with the prior art, the present invention has the following beneficial effects: 1. By deploying a synchronous data acquisition card to acquire the current signal of the current transformer under transient or steady-state conditions, a mathematical model containing sinusoidal components and exponentially decaying aperiodic components is established. The nonlinear least squares method is used to fit and remove the aperiodic components. Then, noise reduction is achieved by sym8 wavelet 5-level decomposition and Stan unbiased risk estimation threshold processing. This solves the problem that traditional processing methods do not specifically handle aperiodic components and that signal interference affects the accuracy of feature extraction, thereby improving the purity of current waveform data.

[0038] 2. The measurement error datasets of the two circuits are obtained by comparing the magnetic field sensor part with a standard magnetic field sensor and the secondary winding part with a standard current transformer. Then, the error data is supplemented by bicubic interpolation, and the error function model is obtained by fitting the least squares method. This solves the problem that the traditional processing method relies on a single standard transformer for comparison and cannot be adapted to the dual-circuit structure, making the error assessment more comprehensive.

[0039] 3. Chirp-Z transform is used to refine the fundamental and harmonic spectra, accurately extracting the amplitude and phase parameters of the fundamental and harmonics. The ratio difference compensation vector is calculated based on the harmonic flux compensation model to eliminate the additional errors introduced by the harmonic components. At the same time, the frequency domain error data is inversely FFT transformed to the time domain and weighted and synthesized according to the proportion of the dual-loop output signal. This solves the shortcomings of traditional processing methods, such as insufficient resolution of spectrum analysis and relying solely on the fundamental wave to calculate the ratio difference, and achieves accurate measurement and comprehensive evaluation of the ratio difference. Attached Figure Description

[0040] Figure 1 This is a flowchart of the steps of a data processing method based on the ratio difference of current transformers according to the present invention.

[0041] Figure 2 This is a schematic diagram of a data processing system based on the ratio difference of current transformers according to the present invention. Detailed Implementation

[0042] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate the technical solutions of the present invention, but should not be used to limit the scope of protection of the present invention.

[0043] Example: As attached Figure 1As shown: This invention provides a data processing method based on the ratio difference of a current transformer, which includes the following steps: S1: Connect the acquisition card to the primary and secondary current detection terminals of the current transformer to obtain the current waveform dataset.

[0044] In this embodiment, a synchronous data acquisition card is deployed, and its signal input ports are connected to the primary and secondary current detection terminals of the current transformer, respectively. The sampling frequency is set according to the rated current parameters of the current transformer and the power grid frequency parameters. The acquisition program is then started to synchronously acquire the primary and secondary current signals of the transformer under transient or steady-state conditions. Signal data is continuously acquired for several power frequency cycles to ensure that the acquired signals cover the complete current change cycle, thus obtaining a dataset of original primary and secondary current waveforms. Based on this dataset, a current mathematical model is established, including sinusoidal components and exponentially decaying aperiodic components. This model can simultaneously characterize the periodic AC component and the transient DC decay component in the current signal. The parameters in the model are iteratively solved using a nonlinear least squares method. Through multiple iterations, the sum of squared errors between the current data output by the model and the actual acquired original current data is minimized, thus obtaining the optimal aperiodic component. The fitting results are used to subtract the optimally fitted aperiodic component from the original current waveform data point by point to eliminate the interference of transient DC attenuation components on subsequent analysis, thereby obtaining a corrected current waveform dataset. Based on the corrected current waveform dataset, a 5-level wavelet decomposition operation is performed on the data using the sym8 wavelet basis function. After decomposition, the approximation coefficients and detail coefficients corresponding to each level are obtained. The approximation coefficients correspond to the low-frequency principal components of the signal, and the detail coefficients correspond to the high-frequency noise components of different frequency bands. Based on the Stan unbiased risk estimation criterion, the optimal thresholds for the detail coefficients of levels 1 to 5 are adaptively calculated to avoid the limitations of manually setting thresholds. A soft thresholding function is used to process the detail coefficients of each level, setting the detail coefficients with absolute values ​​less than the optimal threshold to zero and shrinking the detail coefficients with absolute values ​​greater than the optimal threshold towards zero, thus preserving the effective high-frequency features in the signal. The processed detail coefficients of each level and the approximation coefficients of level 5 are then reconstructed using wavelet to restore the denoised current signal and obtain the current waveform dataset.

[0045] S2: Perform dual electromagnetic loop calibration measurement based on current waveform dataset to obtain measurement error datasets for the two electromagnetic loops. Perform bicubic interpolation and fitting based on the measurement error datasets for the two electromagnetic loops to obtain error function model dataset.

[0046] In this embodiment, based on the current waveform dataset, the magnetic field sensor section and the secondary winding section of the magnetic valve-type current transformer are selected respectively. The signal output terminal of the magnetic field sensor section and the signal output terminal of the standard magnetic field sensor are connected to the same data recording device. DC signals and AC signals of different frequencies are applied to the magnetic valve-type current transformer, and the output signal amplitude and phase information of the magnetic field sensor section and the standard magnetic field sensor under the corresponding signal input are recorded respectively. The two sets of data are compared and calculated to obtain the measurement error of the magnetic field sensor section under different signal types. The secondary winding section and the standard current transformer are connected to the same test circuit, and the current frequency of the test circuit is adjusted. Under different current frequency conditions, the output current amplitude and phase information of the secondary winding section and the standard current transformer are recorded respectively. The difference between the two sets of data is compared to obtain the amplitude error and phase angle error of the secondary winding section. The error data of the magnetic field sensor section and the secondary winding section are integrated to obtain the measurement error dataset of the two electromagnetic circuits. Based on the measurement of the two electromagnetic circuits... The error dataset is generated by selecting 16 valid measurement points around an arbitrary target point in the frequency-amplitude error plane. The distance from each measurement point to the target point is calculated in both the frequency and amplitude dimensions. Weighting coefficients for each measurement point are determined using bicubic interpolation basis functions. The error measurement value of each measurement point is multiplied by its corresponding weighting coefficient, and all products are summed to obtain the estimated error value for the target point. This method is used to supplement data in areas where no actual measurements were taken within the frequency and amplitude range, resulting in an interpolated error supplement dataset. Based on this supplement dataset, a correlation function between the error value and the frequency and amplitude is constructed. The frequency and amplitude values ​​in the supplement dataset are used as input variables, and the corresponding error values ​​are used as output variables. The least squares method is used to iteratively solve the coefficients of the correlation function, minimizing the sum of squared deviations between the calculated function value and the actual error value. Finally, an error function model dataset that characterizes the variation of error with frequency and amplitude is obtained.

[0047] S3: Based on the current waveform dataset, refine the data to obtain the amplitude and phase angle datasets of the primary and secondary fundamental waves and harmonics. Perform FFT transformation on the current waveform dataset and fit it with the error function model dataset to obtain signal error datasets of different frequencies in the frequency domain of the two loops.

[0048] In this embodiment, based on the current waveform dataset, a narrowband analysis frequency range centered on each nominal frequency is set for the fundamental wave and the nominal frequencies of the 3rd, 5th, and 7th harmonics in the current signal. This range can cover the fluctuation range of the corresponding frequency. Chirp-Z transform is performed within the set narrowband analysis frequency range to achieve spectrum refinement, obtaining high-resolution discrete spectrum data. Then, secondary interpolation calculations are performed on the peak points in the discrete spectrum and their two adjacent data points to accurately determine the amplitude and phase parameters corresponding to each frequency component. The amplitude and phase information of the primary and secondary fundamental waves and each harmonic are integrated to obtain the amplitude and phase angle datasets of the primary and secondary fundamental waves and harmonics. Based on the current waveform dataset... The waveform dataset is used to extract current signal data corresponding to the magnetic field sensor and secondary winding parts, respectively. FFT transformation is performed on the two sets of signal data to convert the time-domain signals into frequency-domain signals, resulting in amplitude-frequency datasets corresponding to the output signals of the two electromagnetic loops. This dataset contains signal amplitude information corresponding to different frequency points. Based on the amplitude-frequency datasets of the output signals of the two electromagnetic loops and the error function model dataset, the amplitude data and corresponding frequency values ​​of each frequency point in the amplitude-frequency dataset are substituted into the error function model one by one. The signal error value corresponding to each frequency point is calculated by the model. The error calculation results of all frequency points are integrated to obtain signal error datasets of different frequencies in the frequency domain of the two loops.

[0049] S4: Calculate the ratio difference based on the primary and secondary fundamental wave and harmonic amplitude and phase angle datasets respectively, and input the preset harmonic flux compensation model to obtain the ratio difference dataset and the ratio difference compensation vector dataset.

[0050] In this embodiment, based on the primary and secondary fundamental and harmonic amplitude phase angle datasets, the primary and secondary fundamental current amplitudes, refined by Chirp-Z transform, are extracted. Simultaneously, the nominal transformation ratio of the magnetic valve-type current transformer, calibrated at the factory, is retrieved from the system's stored transformer parameter library. The extracted secondary fundamental current amplitude is multiplied by the nominal transformation ratio to obtain the equivalent current amplitude of the secondary current referred to the primary side. This equivalent current amplitude is subtracted from the measured primary fundamental current amplitude to obtain the deviation between the primary and secondary current amplitudes. This deviation is then divided by the measured primary fundamental current amplitude to obtain the relative deviation ratio of the transformer ratio. The relative deviation ratio is then converted... The percentage results under different operating conditions are integrated to obtain the ratio difference dataset. Based on the fundamental and harmonic amplitude and phase angle datasets of the first and second sides, the amplitude and phase angle parameters of the second-side fundamental wave and the third, fifth, and seventh harmonics are extracted. These parameters are input into a preset harmonic flux compensation model. The proportional coefficient in the harmonic flux compensation model is calibrated by a multiple linear regression algorithm. The calibration process involves applying a standard current containing the fundamental wave and a set harmonic component to the current transformer under test, simultaneously measuring the actual error and fitting the correspondence between harmonic parameters and error. The additional error compensation amount introduced by the harmonic component is calculated by the model. The compensation amount data corresponding to different frequency harmonics are integrated to obtain the ratio difference compensation vector dataset.

[0051] S5: Based on the ratio difference dataset and the ratio difference compensation vector dataset, perform vector summation, convert the signal error datasets of different frequencies in the frequency domain of the two circuits back to the time domain, and perform weighted synthesis with the vector summation result to obtain the current transformer ratio difference dataset.

[0052] In this embodiment, based on the ratio difference dataset and the ratio difference compensation vector dataset, the initial ratio difference in the ratio difference dataset is taken as the real part, and the ratio difference compensation vector is converted into the corresponding parameters as the imaginary part, thus forming the initial ratio difference complex vector and the ratio difference compensation complex vector, respectively. Vector summation is performed on the two complex vectors to cancel the additional error introduced by harmonic components. The calculation results under all operating conditions are integrated to obtain the compensated initial ratio difference dataset. Based on the signal error datasets of different frequencies in the frequency domain of the two loops, inverse FFT transformation is performed on the frequency domain error data of the magnetic field sensor part and the secondary winding part, respectively, to convert the frequency domain error data back to the time domain dimension, resulting in... Two sets of data that reflect the variation of error over time are obtained and integrated to form time-domain signal error datasets for two electromagnetic circuits. Error synthesis is performed based on the initial ratio difference dataset after compensation and the time-domain signal error datasets of the two electromagnetic circuits. The proportion of the output signals of the magnetic field sensor and the secondary winding in the final measurement signal is retrieved from the system parameter library. According to this proportion, corresponding weight coefficients are assigned to the time-domain signal error datasets of the two electromagnetic circuits. The weighted time-domain signal error data is fused with the initial ratio difference data after compensation, and the results of the effects of all error influencing factors are integrated to obtain the current transformer ratio difference dataset.

[0053] Please see as follows Figure 2 As shown, the present invention also provides a data processing system based on the ratio difference of a current transformer, including: a current waveform data acquisition module, including an acquisition card connected to the primary and secondary current detection terminals of the current transformer, for acquiring current waveform datasets.

[0054] The dual electromagnetic loop calibration and fitting module is connected to the current waveform data acquisition module. It is used to perform dual electromagnetic loop calibration measurement and fitting based on the current waveform dataset to obtain the error function model dataset.

[0055] The frequency domain error analysis module is connected to the current waveform data acquisition module and the dual electromagnetic loop calibration and fitting module. It is used to refine the current waveform dataset, and at the same time, it performs FFT transformation based on the current waveform dataset and fits it with the error function model dataset to obtain signal error datasets of different frequencies in the frequency domain of the two loops.

[0056] The compensation module, connected to the frequency domain error analysis module, includes a preset harmonic flux compensation model for acquiring the ratio difference compensation vector dataset;

[0057] The error synthesis module, connected to the compensation module and the frequency domain error analysis module, is used to perform vector summation on the ratio difference dataset and the ratio difference compensation vector dataset, convert the signal error datasets of different frequencies in the frequency domain of the two circuits back to the time domain, and perform weighted synthesis with the vector summation result to obtain the current transformer ratio difference dataset.

[0058] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A data processing method based on the ratio difference of current transformers, characterized in that, It includes the following steps: S1: Connect the data acquisition card to the primary and secondary current detection terminals of the current transformer to obtain the current waveform dataset; S2: Perform dual electromagnetic loop calibration measurement based on current waveform dataset to obtain measurement error datasets for the two electromagnetic loops. Perform bicubic interpolation and fitting based on the measurement error datasets for the two electromagnetic loops to obtain error function model dataset. S3: Based on the current waveform dataset, refine the fundamental and harmonic amplitude and phase angle datasets of the first and second sides, perform FFT transformation on the current waveform dataset and fit it with the error function model dataset to obtain signal error datasets of different frequencies in the frequency domain of the two loops. S4: Calculate the ratio difference based on the amplitude and phase angle datasets of the primary and secondary fundamental waves and harmonics respectively, and input the preset harmonic flux compensation model to obtain the ratio difference dataset and the ratio difference compensation vector dataset. S5: Based on the ratio difference dataset and the ratio difference compensation vector dataset, perform vector summation, convert the signal error datasets of different frequencies in the frequency domain of the two circuits back to the time domain, and perform weighted synthesis with the vector summation result to obtain the current transformer ratio difference dataset.

2. The data processing method based on the ratio difference of current transformers according to claim 1, characterized in that, The step of connecting the acquisition card to the primary and secondary current detection terminals of the current transformer to obtain the current waveform dataset includes: Deploy a synchronous data acquisition card, connect the acquisition card to the primary and secondary current detection terminals of the current transformer, set the sampling frequency, synchronously acquire the current signal of the transformer under transient or steady-state operating conditions, continuously acquire several power frequency cycles, and obtain the original current waveform dataset of the primary and secondary sides. A current mathematical model containing sinusoidal components and exponentially decaying aperiodic components is established based on the original current waveform dataset of the primary and secondary sides. The model parameters are solved iteratively using the nonlinear least squares method to obtain the optimal fitting result of the aperiodic component. This component is then subtracted from the original data to obtain the corrected current waveform dataset. The corrected current waveform dataset is then decomposed to obtain the current waveform dataset.

3. The data processing method based on the ratio difference of current transformers according to claim 2, characterized in that, The process of decomposing the current waveform dataset based on the corrected current waveform dataset to obtain the current waveform dataset includes: Based on the corrected current waveform dataset, the data is decomposed into 5 levels using the sym8 wavelet basis function to obtain the approximation coefficients and detail coefficients of each level. Based on the Stan unbiased risk estimation criterion, the optimal threshold is adaptively calculated for the detail coefficients of levels 1 to 5. The detail coefficients are processed by a soft threshold function, and the processed detail coefficients are reconstructed with the approximation coefficients of level 5 to obtain the current waveform dataset.

4. The data processing method based on the ratio difference of current transformers according to claim 1, characterized in that, The dual electromagnetic loop calibration measurement based on the current waveform dataset obtains measurement error datasets for the two electromagnetic loops. Based on these datasets, bicubic interpolation and fitting are performed to obtain an error function model dataset, including: Based on the current waveform dataset, the magnetic field sensor section and the secondary winding section of the magnetic valve current transformer are selected respectively. The magnetic field sensor section is compared with the standard magnetic field sensor to measure the output signals of DC and AC signals of different frequencies. The secondary winding section is compared with the standard current transformer to measure the output signals under different current frequencies, and the measurement error datasets of the two electromagnetic circuits are obtained. Based on the measurement error dataset of two electromagnetic loops, 16 measurement points around the target point are selected, the weight coefficient corresponding to the distance from each point to the target point is calculated, the error measurement value of each point is multiplied by the corresponding coefficient and then summed to supplement the error data in the frequency and amplitude range, and the interpolated error supplement dataset is obtained. The error function model dataset is obtained by least squares fitting based on the interpolated error supplement dataset.

5. The data processing method based on the ratio difference of current transformers according to claim 1, characterized in that, The process involves refining the current waveform dataset to obtain the amplitude and phase angle datasets of the primary and secondary fundamental waves and harmonics, performing an FFT transformation on the current waveform dataset and fitting it with the error function model dataset to obtain signal error datasets of different frequencies in the frequency domain of the two loops, including: Based on the current waveform dataset, for the nominal frequencies of the fundamental wave and the 3rd, 5th and 7th harmonics, a narrow-band analysis frequency range centered on the nominal frequency is set. Chirp Z transform is performed within the narrow-band range to refine the spectrum. Secondary interpolation is performed on the peak points of the discrete spectrum and their adjacent points to obtain the amplitude and phase angle datasets of the fundamental wave and harmonics on the first and second sides. Based on the current waveform dataset, the output signals of the magnetic field sensor section and the secondary winding section are converted into the frequency domain to obtain the amplitude-frequency datasets of the output signals of the two electromagnetic circuits. Based on the amplitude-frequency datasets of the output signals of the two electromagnetic circuits and the error function model dataset, the amplitude-frequency data are substituted into the error function model to calculate the signal error at different frequencies, and obtain the signal error datasets of the two circuits at different frequencies in the frequency domain.

6. The data processing method based on the ratio difference of current transformers according to claim 1, characterized in that, The process of calculating the ratio difference based on the amplitude and phase angle datasets of the primary and secondary fundamental waves and harmonics, and inputting them into a preset harmonic flux linkage compensation model, to obtain the ratio difference dataset and the ratio difference compensation vector dataset includes: Based on the primary and secondary fundamental and harmonic amplitude and phase angle datasets, the primary and secondary fundamental current amplitudes obtained by ChirpZ transform are extracted. At the same time, the nominal ratio calibrated at the time of manufacture of the magnetic valve current transformer is retrieved. The secondary fundamental current amplitude is multiplied by the nominal ratio to obtain the equivalent current amplitude of the secondary current converted to the primary side. The current amplitude deviation is obtained by subtracting the primary fundamental current amplitude from the equivalent current amplitude. The relative deviation ratio of the transformer ratio is obtained by dividing the current amplitude deviation by the primary fundamental current amplitude. Thus, the ratio difference dataset is obtained. Based on the amplitude and phase angle datasets of the first and second fundamental waves and harmonics, the amplitudes and phase angles of the second fundamental wave and the third, fifth and seventh harmonics are input into a preset harmonic flux compensation model. The proportional coefficients in the harmonic flux compensation model are calibrated by a multiple linear regression algorithm, and the proportional difference compensation vector dataset is calculated.

7. The data processing method based on the ratio difference of current transformers according to claim 1, characterized in that, The process of vector summation based on the ratio difference dataset and the ratio difference compensation vector dataset, converting the signal error datasets of different frequencies in the frequency domain of the two circuits back to the time domain, and weighting and synthesizing them with the vector summation result to obtain the current transformer ratio difference dataset includes: Based on the ratio difference dataset and the ratio difference compensation vector dataset, the initial ratio difference and the ratio difference compensation vector are vector summed to eliminate the additional error introduced by the harmonic components and obtain the compensated initial ratio difference dataset. Inverse FFT transformation is performed on the signal error datasets of the two loops at different frequencies in the frequency domain to convert the frequency domain error data back to the time domain and obtain the time domain signal error datasets of the two electromagnetic loops. Error synthesis is performed based on the initial ratio difference dataset after compensation and the time-domain signal error dataset of the two electromagnetic circuits. According to the proportion of the output signals of the magnetic field sensor section and the secondary winding section in the final measurement signal, the two types of error data are weighted and synthesized to obtain the current transformer ratio difference dataset.

8. A data processing system based on the ratio difference of a current transformer, characterized in that, include: The current waveform data acquisition module includes acquisition cards connected to the primary and secondary current detection terminals of the current transformer, used to acquire current waveform datasets. The dual electromagnetic loop calibration and fitting module is connected to the current waveform data acquisition module. It is used to perform dual electromagnetic loop calibration measurement and fitting based on the current waveform dataset to obtain the error function model dataset. The frequency domain error analysis module is connected to the current waveform data acquisition module and the dual electromagnetic loop calibration and fitting module. It is used to refine the current waveform dataset, and at the same time, it performs FFT transformation based on the current waveform dataset and fits it with the error function model dataset to obtain signal error datasets of different frequencies in the frequency domain of the two loops. The compensation module, connected to the frequency domain error analysis module, includes a preset harmonic flux compensation model for acquiring the ratio difference compensation vector dataset; The error synthesis module, connected to the compensation module and the frequency domain error analysis module, is used to perform vector summation on the ratio difference dataset and the ratio difference compensation vector dataset, convert the signal error datasets of different frequencies in the frequency domain of the two circuits back to the time domain, and perform weighted synthesis with the vector summation result to obtain the current transformer ratio difference dataset.

Images