Evaluation method and system for online monitoring device of voltage transformer

Abstract

The invention discloses an evaluation method and system for a voltage transformer on-line monitoring device, and the method comprises the steps: completing the error evaluation of each channel of the voltage transformer on-line monitoring device under the condition of not adding a ratio difference and a phase difference; partial channels of the voltage transformer on-line monitoring device are selected for difference adding, and error evaluation of each channel of the voltage transformer on-line monitoring device is completed; and calculating a ratio difference accuracy index and a phase difference accuracy index by using the error evaluation results before and after difference addition, and if the ratio difference accuracy index is less than or equal to a first preset value and the phase difference accuracy index is less than or equal to a second preset value, determining that the voltage transformer online monitoring device is normally evaluated. Otherwise, the voltage transformer on-line monitoring device is evaluated to be abnormal and needs to be overhauled; the method has the advantages that whether the state of the online monitoring device of the voltage transformer is normal or not is evaluated, so that timely maintenance of the online monitoring device of the voltage transformer is facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of online monitoring of electric power metering, and in particular to an evaluation method and system for an online monitoring device of a voltage transformer. Background Art

[0002] The voltage transformer online monitoring device is installed at the substation site. By collecting the secondary voltage information of the voltage transformer and based on the online evaluation algorithm, the voltage transformer error measurement status is evaluated. However, during the long-term energized operation, the online monitoring device may experience performance drift, resulting in performance degradation. In order to ensure the stable operation of the voltage transformer online monitoring device in operation, it is necessary to carry out on-site verification of the voltage transformer online monitoring device to evaluate whether the status of the voltage transformer online monitoring device is normal.

[0003] Chinese patent publication number CN110095747A discloses an online monitoring method and system for voltage and current sensors for distribution networks, wherein the method comprises: establishing a reference channel, including: converting the secondary output signal of an electrical current and voltage transformer through a signal conversion device and then outputting it to an analog-to-digital converter; establishing a measured channel, including: receiving the secondary side output signal of the measured current and voltage transformer through a front circuit, and outputting the received secondary side output signal of the measured current and voltage transformer to the analog-to-digital converter; converting the analog signals output by the reference channel and the measured channel into digital signals through the analog-to-digital converter, and sending the converted digital signals of the reference channel and the measured channel to an error calculation module; calculating the ratio error and phase error of the converted digital signals of the reference channel and the measured channel through the error calculation module; analyzing the error change trend of the measured current and voltage transformer through a data analysis module, and judging the state of the reference channel. This patent application calculates the error between the reference channel and the measured channel, as well as the error between the measured channels, and can determine whether the reference channel is in a normal state. However, its solution is relatively vague and does not disclose specific rules for determining whether the state is normal or not. As a result, it is impossible to accurately evaluate whether the state of the voltage transformer online monitoring device is normal. Summary of the invention

[0004] The technical problem to be solved by the present invention is that the prior art cannot evaluate whether the state of an online monitoring device of a voltage transformer is normal, and thus cannot timely repair the online monitoring device of the voltage transformer.

[0005] The present invention solves the above technical problems through the following technical means: an evaluation method for an online monitoring device of a voltage transformer, comprising: S1. Without adding ratio difference and phase difference, a voltage signal is input to each channel of the voltage transformer online monitoring device to complete error evaluation of each channel of the voltage transformer online monitoring device; S2. When the input voltage signal remains unchanged, select some channels of the voltage transformer online monitoring device in turn to perform differential addition, so as to complete the error evaluation of each channel of the voltage transformer online monitoring device; the differential addition refers to adding ratio difference and phase difference; S3. Calculate the ratio difference accuracy index and the phase difference accuracy index using the error evaluation results before and after the difference addition. If the ratio difference accuracy index is less than or equal to the first preset value and the phase difference accuracy index is less than or equal to the second preset value, the voltage transformer online monitoring device is evaluated normally. Otherwise, the voltage transformer online monitoring device is evaluated abnormally and requires maintenance.

[0006] Furthermore, in said S2, some channels of the voltage transformer online monitoring device are selected to perform analog-to-digital conversion on the input voltage signal before adding the difference.

[0007] Furthermore, the method of adding phase difference to some channels of the voltage transformer online monitoring device in S2 is: S201, preprocessing the original signal to obtain a reference wave, and dividing the original signal into a plurality of sub-waves; S202, calculating similarity features between each wavelet and the reference wave, and calculating similarity between each wavelet and the reference wave according to the similarity features, wherein the similarity features include distance features, shape features, and data features; S203, selecting a wavelet whose similarity exceeds a set threshold as a candidate wavelet based on the similarity between each wavelet and the reference wave; selecting a band on the left side of the candidate wavelet as a first candidate extended wavelet, and selecting the first candidate extended wavelet according to the predicted maximum, minimum, and slope of the left extended wave of the original signal and the maximum, minimum, and slope of each first candidate extended wavelet, to obtain the left extended wave of the original signal; S204, selecting the waveband on the right side of the candidate wavelet as the second candidate extended wave, and selecting the second candidate extended wave according to the predicted maximum, minimum, slope of the right side extended wave of the original signal and the maximum, minimum, and slope of each second candidate extended wave, to obtain the right side extended wave of the original signal; S205, based on the left side extended wave of the original signal and the right side extended wave of the original signal, the original signal is extended to the left and right respectively, and the extended waveform is subjected to windowed Hilbert transform to realize the phase addition and difference of the corresponding channel of the voltage transformer online monitoring device.

[0008] Furthermore, the S201 includes: Suppose the original signal is e(N), and the corresponding left endpoint is , past Draw a line parallel to the horizontal axis, and its intersection points with e(N) are , , is the total number of intersection points; As the starting point, the intercept length is The signal is used as the reference wave , when selected Contains a maximum point, a minimum point and a zero crossing point. is the number of sampling points of the reference wave, The reference wave The amplitude of the sampling points; As the starting point, the original signal is divided into K sub-waves , ,... , each sub-wave contains a maximum point, a minimum point and a zero-crossing point, and the number of sampling points of each sub-wave is recorded as .

[0009] Furthermore, the process of S202 is: For Sub-wave , which is consistent with the reference wave The distance feature is

[0010] in, For the The number of sampling points of a wavelet, For the The sub-wave The amplitude of the sampling points, Indicates the calculation process from point (1,1) to point The cumulative regularization distance is calculated as follows:

[0011] In the formula, Indicates Sub-wave With Sub-wave The Euclidean distance between Before calculating the shape features between different waveforms, the waveform data of each sub-wave needs to be interpolated or sampled according to the reference wave to adjust the number of sampling points of the sub-wave to be consistent with that of the reference wave. The interpolation or sampling ratio of the wavelets is: ; Compute shape features:

[0012] in, is the sampling point number and =1,2,..., , , They are the reference wave, The sub-wave The amplitude of the sampling points, is an exponential function with base e; Calculate data features:

[0013] in, , They are the reference wave, The mean amplitude of the sub-waves at the sampling points; The similarity between each wavelet and the reference wave is calculated based on the similarity characteristics. The similarity calculation formula between a sub-wave and the reference wave is:

[0014] in, They are the set first weight, second weight and third weight respectively.

[0015] Furthermore, the process of S203 is: Based on the similarity between each sub-wave and the reference wave, Select the corresponding wavelet as the candidate wavelet, where For the set threshold, select the band on the left side of the candidate sub-wave as the first candidate extension wave:

[0016] in, represents the total number of first candidate extension waves, is the u-th first candidate extension wave, , Indicates the Uth first candidate extension wave The amplitude of the sampling points; The horizontal and vertical coordinates of the maximum value of the left extension wave of the original signal are predicted ( ), the horizontal and vertical coordinates of the minimum value ( ), slope ; At the same time, the maximum, minimum, and slope of each first candidate extension wave are calculated, and the left extension wave of the original signal is selected from the first candidate extension wave based on the following formula:

[0017]

[0018] in, is the uth wave to be selected, Indicate point With point The Euclidean distance of are the ordinates of the maximum and minimum of the u-th first candidate extended wave, respectively. is the slope of the u-th first candidate extension wave, For the fourth and fourth and fifth weights, It is the left extension wave of the original signal.

[0019] Furthermore, the prediction obtains the horizontal and vertical coordinates of the maximum value of the left extension wave of the original signal ( ), the horizontal and vertical coordinates of the minimum value ( ), slope The process is: Assume the original signal is (N), build a maximum value data set , Minimum Dataset , slope set :

[0020]

[0021]

[0022] in, is the number of maximum value data sets, minimum value data sets and slope concentrated data sets of the original signal, Represents the horizontal and vertical coordinate values ​​of the nth maximum point of the original signal from left to right, Represents the horizontal and vertical coordinate values ​​of the nth minimum point of the original signal from left to right, It represents the nth slope of the original signal from left to right. ; The LSTM model is used to predict the maximum point, minimum point and slope of the original signal, and based on the above data set , , The loss function is constructed based on the prediction results of the LSTM model, and the LSTM model is trained. When the loss function is the smallest, the training is stopped to obtain the trained LSTM model. The loss function is:

[0023] Where j represents the jth training sample, j=1,2,...J; Represents the number of model training samples. The jth training sample includes the horizontal and vertical coordinate values ​​of the jth maximum point of the original signal from left to right. , the horizontal and vertical coordinates of the jth minimum point of the original signal from left to right And the jth slope of the original signal from left to right ; ( ) represents the horizontal and vertical coordinates predicted values ​​of the jth maximum point of the original signal from left to right, ( ) represents the horizontal and vertical coordinates predicted values ​​of the jth minimum point of the original signal from left to right, Represents the predicted value of the jth slope of the original signal from left to right; Use the trained LSTM model to predict the horizontal and vertical coordinates of the maximum value of the left extension wave of the original signal ( ), the horizontal and vertical coordinates of the minimum value ( ), slope .

[0024] Furthermore, the process of S3 is: Before adding the difference, the evaluation result of the voltage transformer online monitoring device on the i-th voltage transformer is: , represents the ratio error evaluation result of the i-th voltage transformer before adding the difference, represents the phase error evaluation result of the i-th voltage transformer before adding the difference; the evaluation result of the voltage transformer online monitoring device on the i-th voltage transformer after adding the difference is , represents the ratio error evaluation result of the i-th voltage transformer after adding the difference, represents the phase error evaluation result of the i-th voltage transformer after adding the difference, then the ratio difference accuracy index of the voltage transformer online monitoring device is:

[0025] in, It is the theoretical difference between the ratio error before adding the difference and the ratio error measured after adding the difference; The phase difference accuracy index of the voltage transformer online monitoring device is:

[0026] in, is the theoretical difference between the phase error before adding the difference and the phase error measured after adding the difference. If , , then the voltage transformer online monitoring device is evaluated to be normal, otherwise the voltage transformer online monitoring device is evaluated to be abnormal and needs to be repaired, among which, Represents a phase angle of 2 minutes.

[0027] The present invention also provides an evaluation system for an online monitoring device of a voltage transformer, comprising: The first evaluation module is used to input voltage signals to each channel of the voltage transformer online monitoring device without adding ratio difference and phase difference, so as to complete error evaluation of each channel of the voltage transformer online monitoring device; The second evaluation module is used to select some channels of the voltage transformer online monitoring device in turn to perform differential addition when the input voltage signal remains unchanged, so as to complete the error evaluation of each channel of the voltage transformer online monitoring device; the differential addition refers to adding ratio difference and phase difference; The accuracy assessment module is used to calculate the ratio difference accuracy index and the phase difference accuracy index using the error assessment results before and after the difference addition. If the ratio difference accuracy index is less than or equal to the first preset value and the phase difference accuracy index is less than or equal to the second preset value, the voltage transformer online monitoring device is evaluated normally; otherwise, the voltage transformer online monitoring device is evaluated abnormally and requires maintenance.

[0028] Furthermore, the second evaluation module selects some channels of the voltage transformer online monitoring device to perform analog-to-digital conversion on the input voltage signal before adding the difference.

[0029] Furthermore, the method for adding phase difference to some channels of the voltage transformer online monitoring device in the second evaluation module is: S201, preprocessing the original signal to obtain a reference wave, and dividing the original signal into a plurality of sub-waves; S202, calculating similarity features between each wavelet and the reference wave, and calculating similarity between each wavelet and the reference wave according to the similarity features, wherein the similarity features include distance features, shape features, and data features; S203, selecting a wavelet whose similarity exceeds a set threshold as a candidate wavelet based on the similarity between each wavelet and the reference wave; selecting a band on the left side of the candidate wavelet as a first candidate extended wavelet, and selecting the first candidate extended wavelet according to the predicted maximum, minimum, and slope of the left extended wave of the original signal and the maximum, minimum, and slope of each first candidate extended wavelet, to obtain the left extended wave of the original signal; S204, selecting the waveband on the right side of the candidate wavelet as the second candidate extended wave, and selecting the second candidate extended wave according to the predicted maximum, minimum, slope of the right side extended wave of the original signal and the maximum, minimum, and slope of each second candidate extended wave, to obtain the right side extended wave of the original signal; S205, based on the left side extended wave of the original signal and the right side extended wave of the original signal, the original signal is extended to the left and right respectively, and the extended waveform is subjected to windowed Hilbert transform to realize the phase addition and difference of the corresponding channel of the voltage transformer online monitoring device.

[0030] Further, the S201 includes: Suppose the original signal is e(N), and the corresponding left endpoint is , past Draw a line parallel to the horizontal axis, and its intersection points with e(N) are , , is the total number of intersection points; As the starting point, the intercept length is The signal is used as the reference wave , when selected Contains a maximum point, a minimum point and a zero crossing point. is the number of sampling points of the reference wave, The reference wave The amplitude of the sampling points; As the starting point, the original signal is divided into K sub-waves , ,... , each sub-wave contains a maximum point, a minimum point and a zero-crossing point, and the number of sampling points of each sub-wave is recorded as .

[0031] Furthermore, the process of S202 is: For Sub-wave , which is consistent with the reference wave The distance feature is

[0032] in, For the The number of sampling points of a wavelet, For the The sub-wave The amplitude of the sampling points, Indicates the calculation process from point (1,1) to point The cumulative regularization distance is calculated as follows:

[0033] In the formula, Indicates Sub-wave With Sub-wave The Euclidean distance between Before calculating the shape features between different waveforms, the waveform data of each sub-wave needs to be interpolated or sampled according to the reference wave to adjust the number of sampling points of the sub-wave to be consistent with that of the reference wave. The interpolation or sampling ratio of the wavelets is: ; Compute shape features:

[0034] in, is the sampling point number and =1,2,..., , , They are the reference wave, The sub-wave The amplitude of the sampling points, is an exponential function with base e; Calculate data features:

[0035] in, , They are the reference wave, The mean amplitude of the sub-waves at the sampling points; The similarity between each wavelet and the reference wave is calculated based on the similarity characteristics. The similarity calculation formula between a sub-wave and the reference wave is:

[0036] in, They are the set first weight, second weight and third weight respectively.

[0037] Furthermore, the process of S203 is: Based on the similarity between each sub-wave and the reference wave, Select the corresponding wavelet as the candidate wavelet, where For the set threshold, select the band on the left side of the candidate sub-wave as the first candidate extension wave:

[0038] in, represents the total number of first candidate extension waves, is the u-th first candidate extension wave, , Indicates the Uth first candidate extension wave The amplitude of the sampling points; The horizontal and vertical coordinates of the maximum value of the left extension wave of the original signal are predicted ( ), the horizontal and vertical coordinates of the minimum value ( ), slope ; At the same time, the maximum, minimum, and slope of each first candidate extension wave are calculated, and the left extension wave of the original signal is selected from the first candidate extension wave based on the following formula:

[0039]

[0040] in, is the uth wave to be selected, Indicate point With point The Euclidean distance of are the ordinates of the maximum and minimum of the u-th first candidate extended wave, respectively. is the slope of the u-th first candidate extension wave, For the fourth and fourth and fifth weights, It is the left extension wave of the original signal.

[0041] Furthermore, the prediction obtains the horizontal and vertical coordinates of the maximum value of the left extension wave of the original signal ( ), the horizontal and vertical coordinates of the minimum value ( ), slope The process is: Assume the original signal is (N), build a maximum value data set , Minimum Dataset , slope set :

[0042]

[0043]

[0044] in, is the number of maximum value data sets, minimum value data sets and slope concentrated data sets of the original signal, Represents the horizontal and vertical coordinate values ​​of the nth maximum point of the original signal from left to right, Represents the horizontal and vertical coordinate values ​​of the nth minimum point of the original signal from left to right, It represents the nth slope of the original signal from left to right. ; The LSTM model is used to predict the maximum point, minimum point and slope of the original signal, and based on the above data set , , The loss function is constructed based on the prediction results of the LSTM model, and the LSTM model is trained. When the loss function is the smallest, the training is stopped to obtain the trained LSTM model. The loss function is:

[0045] Where j represents the jth training sample, j=1,2,...J; Represents the number of model training samples. The jth training sample includes the horizontal and vertical coordinate values ​​of the jth maximum point of the original signal from left to right. , the horizontal and vertical coordinates of the jth minimum point of the original signal from left to right And the jth slope of the original signal from left to right ; ( ) represents the horizontal and vertical coordinates predicted values ​​of the jth maximum point of the original signal from left to right, ( ) represents the horizontal and vertical coordinates predicted values ​​of the jth minimum point of the original signal from left to right, Represents the predicted value of the jth slope of the original signal from left to right; Use the trained LSTM model to predict the horizontal and vertical coordinates of the maximum value of the left extension wave of the original signal ( ), the horizontal and vertical coordinates of the minimum value ( ), slope .

[0046] Furthermore, the accuracy assessment module is also used to: Before adding the difference, the evaluation result of the voltage transformer online monitoring device on the i-th voltage transformer is: , represents the ratio error evaluation result of the i-th voltage transformer before adding the difference, represents the phase error evaluation result of the i-th voltage transformer before adding the difference; the evaluation result of the voltage transformer online monitoring device on the i-th voltage transformer after adding the difference is , represents the ratio error evaluation result of the i-th voltage transformer after adding the difference, represents the phase error evaluation result of the i-th voltage transformer after adding the difference, then the ratio difference accuracy index of the voltage transformer online monitoring device is:

[0047] in, It is the theoretical difference between the ratio error before adding the difference and the ratio error measured after adding the difference; The phase difference accuracy index of the voltage transformer online monitoring device is:

[0048] in, is the theoretical difference between the phase error before adding the difference and the phase error measured after adding the difference. If , , then the voltage transformer online monitoring device is evaluated to be normal, otherwise the voltage transformer online monitoring device is evaluated to be abnormal and needs to be repaired, among which, Represents a phase angle of 2 minutes.

[0049] The advantages of the present invention are: (1) The present invention simulates different voltage transformer operating states by adding differential to some channels of the voltage transformer online monitoring device, and compares the evaluation results before and after adding differential, and uses the error evaluation results before and after adding differential to calculate the ratio difference accuracy index and the phase difference accuracy index. If the ratio difference accuracy index is less than or equal to a first preset value and the phase difference accuracy index is less than or equal to a second preset value, the evaluation of the voltage transformer online monitoring device is normal; otherwise, the evaluation of the voltage transformer online monitoring device is abnormal and needs to be repaired, so that the voltage transformer online monitoring device can be repaired in time, which is of great significance for promoting the intelligent development of the power system and improving the operation level of the power grid.

[0050] (2) The present invention realizes accurate phase addition of some channels of the voltage transformer online monitoring device through the phase difference algorithm based on the combination of global and local features provided by S2, thereby further improving the accuracy of the evaluation result. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A schematic diagram of the connection relationship between a portable traceability platform and a voltage transformer online monitoring device in an evaluation method for a voltage transformer online monitoring device disclosed in an embodiment of the present invention; Figure 2 The present invention discloses a flowchart of an evaluation method for an online monitoring device for a voltage transformer according to an embodiment of the present invention. DETAILED DESCRIPTION

[0052] 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 combination with 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.

[0053] Example 1 like Figure 1 and Figure 2 As shown, the present invention provides an evaluation method for an online monitoring device of a voltage transformer, which is applied to a portable traceability platform. The portable traceability platform supports simultaneous verification of 12 signals of the online monitoring device of the voltage transformer. The portable traceability platform includes a voltage signal source, an A / D conversion unit and a differential addition module. The main traceability principle is: output a stable voltage signal through the voltage signal source, superimpose a digital differential signal through the differential addition module, compare the error evaluation results of the online monitoring device before and after the differential addition, and verify the accuracy of the evaluation results of the online monitoring device of the voltage transformer. The differential indicates a small error, and there is no specific standard. In practical applications, ratio error and phase error are added as needed.

[0054] The voltage signal source is used to output stable voltage signals to 12 channels simultaneously. The amplitude and phase of each analog output signal can be adjusted independently to simulate the normal and abnormal states of each voltage transformer, wherein the number of abnormal voltage transformers in a single test is set to not exceed 1 / 3 of the total number of monitored voltage transformers. The hardware of the analog differential source is complex, easily affected by time drift and temperature drift, and has poor stability, resulting in unreliable traceability results. The digital differential source can overcome these drawbacks of analog differentials, so the present invention uses an A / D conversion unit to convert analog signals into digital signals to facilitate the addition of digital differential signals. In this embodiment, the difference addition module is used to add digital ratio differences and digital phase differences to each channel.

[0055] Continue reading Figure 2 The evaluation method of the voltage transformer online monitoring device provided by the present invention comprises the following steps: S1. Without adding ratio difference and phase difference, the voltage signal source in the portable traceability platform outputs 12 voltage signals to 12 channels of the voltage transformer online monitoring device to complete the error evaluation of each channel of the voltage transformer online monitoring device. The specific process is as follows: 1) A portable traceability platform is used to simulate the secondary output of the voltage transformer at the site. The amplitude and phase of each simulated output signal can be adjusted independently to simulate the normal and abnormal states of each voltage transformer. To ensure the accuracy of the evaluation results of the online monitoring device, the number of abnormal voltage transformers in a single test is set to no more than 1 / 3 of the total number of monitored voltage transformers.

[0056] (1) represents the simulated voltage transformer secondary output voltage set, Represents the simulated secondary output voltage of the i-th voltage transformer, i=1,2,3,...,12.

[0057] 2) The voltage transformer online monitoring device completes the error evaluation of each voltage transformer based on the electrical topological relationship between the secondary voltage and the voltage transformer. The evaluation algorithm can be found in the "Online Operation Calibration Method and Device of Voltage Transformer Based on Virtual Standard" disclosed in Chinese Patent Publication No. CN115932702A. The evaluation results are as follows: (2) (3) represents the ratio error evaluation result set of the voltage transformer before adding the differential voltage. represents the phase error evaluation result set of the voltage transformer before adding the difference, represents the ratio error evaluation result of the i-th voltage transformer before adding the difference, It represents the phase error evaluation result of the i-th voltage transformer before adding the difference.

[0058] S2. When the input voltage signal remains unchanged, select some channels of the voltage transformer online monitoring device to connect to the difference adding module, perform difference adding, and complete the error evaluation of each channel of the voltage transformer online monitoring device; the difference adding refers to adding ratio difference and phase difference; To ensure the accuracy of difference adding, the present invention proposes a phase difference algorithm based on the combination of global and local features to achieve accurate addition of phase difference. Before performing difference adding, deviation analysis is performed first, and the analysis process is as follows: 1) The theoretical difference between the ratio difference before adding the difference and the ratio error measured after adding the difference is: (4) is the standard amplitude error superimposed on the simulated i-th voltage transformer.

[0059] 2) The theoretical difference between the phase error before adding the difference and the phase error measured after adding the difference is: (5) is the standard phase error superimposed on the simulated i-th voltage transformer.

[0060] 3) When the standard amplitude error exists When the error is , according to formula (6), the ratio error deviation caused is: (6) in, is the actual value of the difference between the ratio error before adding the difference and the ratio error measured after adding the difference, Indicates that the standard amplitude error added by the i-th voltage transformer exists Error in size.

[0061] When the standard phase error exists When the error is , according to formula (7), the phase error deviation caused is: (7) in, is the actual value of the difference between the phase error before addition and the phase error measured after addition. Indicates that the standard phase error added by the i-th voltage transformer exists Error in size.

[0062] It can be seen from formula (6) that due to , so the influence of the ratio difference error is greatly reduced; it can be seen from formula (7) that the phase difference error is fully reflected in the measurement result. Therefore, in the portable traceability platform, high-precision addition of phase difference is one of the key issues for reliable traceability. Ratio difference addition is to set a ratio difference, and adjust the amplitude of the measured signal by the ratio of the ratio difference to the amplitude of the test signal to achieve ratio difference addition. The following will not repeat the addition of ratio difference, and focus on the addition of phase difference.

[0063] The addition of phase difference is more complicated and needs to be achieved through phase shifting. Here, the Hilbert transform method is used to shift the phase. During the transformation process, a conjugate signal with a phase difference of π / 2 from the original signal is constructed. The conjugate signal is obtained through "Fourier transform-double-sided spectrum folded into single-sided spectrum-Fourier inverse transform". Since the tracing process is affected by frequency offset, harmonic interference and noise, the full cycle sampling cannot be guaranteed, resulting in frequency leakage in Fourier transform, and the single-sided spectrum Fourier inverse transform cannot offset the error caused by frequency leakage, resulting in "flying" at both ends of the signal, that is, there is an endpoint effect, which affects the phase shift accuracy and must be effectively suppressed. Here, a phase difference algorithm based on the combination of global and local features is proposed, as follows: 1) Based on global feature suppression algorithm Considering the signal change trend as a whole, predict the extreme point of the left extension wave of the original signal. Suppose the original signal is (N), build a maximum value data set , Minimum Dataset , slope set : (8) in, is the number of maximum value data sets, minimum value data sets and slope concentrated data sets of the original signal, Represents the horizontal and vertical coordinate values ​​of the nth maximum point of the original signal from left to right, Represents the horizontal and vertical coordinate values ​​of the nth minimum point of the original signal from left to right, It represents the nth slope of the original signal from left to right. .

[0064] Based on the above dataset , , , use the LSTM model to predict the maximum point, minimum point and slope respectively: (9) in,( ) is the horizontal and vertical coordinates of the maximum value of the left extension wave of the predicted original signal, ( ) is the horizontal and vertical coordinates of the minimum value of the left extension wave of the predicted original signal, is the slope of the left extended wave of the predicted original signal.

[0065] The LSTM model is used to predict the maximum point, minimum point and slope of the original signal, and based on the above data set , , The loss function is constructed based on the prediction results of the LSTM model, and the LSTM model is trained. When the loss function is minimized, the training is stopped to obtain the trained LSTM model. The loss function is defined as: (10) Where j represents the jth training sample, j=1,2,...J; Represents the number of model training samples. The jth training sample includes the horizontal and vertical coordinate values ​​of the jth maximum point of the original signal from left to right. , the horizontal and vertical coordinates of the jth minimum point of the original signal from left to right And the jth slope of the original signal from left to right ; ( ) represents the horizontal and vertical coordinates predicted values ​​of the jth maximum point of the original signal from left to right, ( ) represents the horizontal and vertical coordinates predicted values ​​of the jth minimum point of the original signal from left to right, Represents the j-th slope prediction value of the original signal from left to right.

[0066] The input of the LSTM model is the dataset , , , the output is the maximum, minimum and slope of the predicted original signal. For details, refer to formula (9). In order to improve the accuracy of the prediction results, the loss function is set to train the LSTM model. The trained LSTM model can directly output accurate prediction results based on the input data set. When predicting the extreme value of the left extension wave of the original signal, only the set of discrete data points of the left extension wave of the original signal needs to be input. Based on the trained LSTM model, the maximum value of the left extension wave of the original signal is predicted respectively ( ), minimum value ( ), slope . Similarly, the extreme point and slope of the right extended wave can be predicted. The main principle is to predict the maximum value, minimum value and slope value of the left extended wave of the original signal according to the maximum, minimum and slope change trends of the original signal waveform. Then, because there are many alternative extended waves, the most suitable waveform is selected as the extended wave by the approximation between the selected waveform and the predicted waveform. For details, please refer to the following formulas (18) and (19).

[0067] 2) Suppression algorithm based on local features Suppose the original signal is e(N), and the corresponding left endpoint is , past Draw a line parallel to the horizontal axis, and its intersection points with e(N) are , , is the total number of intersection points; As the starting point, the intercept length is The signal is used as the reference wave , when selected Contains a maximum point, a minimum point and a zero crossing point. is the number of sampling points of the reference wave, The reference wave The amplitude of the sampling points; As the starting point, the original signal is divided into K sub-waves , ,... , each sub-wave contains a maximum point, a minimum point and a zero-crossing point. Due to the influence of sampling, transformation and other factors during the signal transformation process, the number of sampling points of each sub-wave and the reference wave may be different, which are recorded as .

[0068] ① Calculate the similarity features between each sub-wave and the reference wave: distance feature, shape feature, and data feature.

[0069] A. Distance Features The dynamic time warping algorithm is used to calculate the distance between the reference waveform and each sub-wave. Sub-wave For example, , , For the The number of sampling points of a wavelet, For the The sub-wave The amplitude of each sampling point.

[0070] For the sequence , , the dynamic time warping distance between them is calculated as follows: (11) (12) (13) In the formula, Indicates the calculation process from point (1,1) to point The cumulative regularization distance of express The distance between two points; .

[0071] Indicates the calculation process from point (1,1) to point The cumulative regularization distance of , Indicates Sub-wave With Sub-wave The Euclidean distance between . Indicates Sub-wave With reference wave distance characteristics.

[0072] B. Shape characteristics Before calculating the shape features between different waveforms, it is necessary to interpolate or sample the waveform data of each sub-wave according to the reference wave to adjust the number of sampling points of the sub-wave to be consistent with that of the reference wave. The interpolation or sampling ratio of the wavelets is: .

[0073] Compute shape features: (14) in, is the sampling point number and =1,2,..., , , They are the reference wave, The sub-wave The amplitude of the sampling points, is an exponential function with base e.

[0074] C. Data characteristics: (15) in, , are the mean amplitudes of the reference wave and the kth sub-wave at the sampling points respectively.

[0075] The above indicators are combined to calculate the similarity between each sub-wave and the reference wave. The similarity calculation formula between a sub-wave and the reference wave is: (16) in, They are the set first weight, second weight and third weight respectively.

[0076] 3) Based on the similarity between each sub-wave and the reference wave, Select the wavelet with the largest similarity to the reference wave as the candidate wavelet, where is the set threshold. Select the band on the left side of the candidate sub-wave as the first candidate extension wave: (17) in, represents the total number of first candidate extension waves, is the u-th first candidate extension wave, , Indicates the Uth first candidate extension wave The amplitude of the sampling points.

[0077] 4) Based on the LSTM model of the global features above, the horizontal and vertical coordinates of the maximum value of the left extension wave of the original signal are obtained ( ), the horizontal and vertical coordinates of the minimum value ( ), slope ; At the same time, the maximum, minimum, and slope of each first candidate extension wave are calculated, and the left extension wave of the original signal is selected from the first candidate extension wave based on the following formula: (18) (19) in, is the uth wave to be selected, Indicate point With point The Euclidean distance of are the ordinates of the maximum and minimum of the u-th first candidate extended wave, respectively. is the slope of the u-th first candidate extension wave, For the fourth and fourth and fifth weights, is the left extension wave of the original signal. Similarly, the right extension wave of the original signal can be obtained.

[0078] 5) Based on the left extended wave of the original signal and the right extended wave of the original signal, the original signal is extended to the left and right respectively, and the extended waveform is subjected to windowed Hilbert transform to realize the phase addition and difference of the corresponding channel of the voltage transformer online monitoring device. Windowed Hilbert transform refers to convolving the extended waveform with a window function to obtain the signal g(t), and then performing Hilbert transform on the signal g(t) to obtain the signal h(t). In this embodiment, the window function is a 4th order 3-term Nuttall window.

[0079] S3. Calculate the ratio difference accuracy index and the phase difference accuracy index using the error evaluation results before and after the difference addition. If the ratio difference accuracy index is less than or equal to the first preset value and the phase difference accuracy index is less than or equal to the second preset value, the voltage transformer online monitoring device is evaluated normally. Otherwise, the voltage transformer online monitoring device is evaluated abnormally and needs to be repaired. The specific process is: Before adding the difference, the evaluation result of the voltage transformer online monitoring device on the i-th voltage transformer is: , represents the ratio error evaluation result of the i-th voltage transformer before adding the difference, represents the phase error evaluation result of the i-th voltage transformer before adding the difference; the evaluation result of the voltage transformer online monitoring device on the i-th voltage transformer after adding the difference is , represents the ratio error evaluation result of the i-th voltage transformer after adding the difference, represents the phase error evaluation result of the i-th voltage transformer after adding the difference, then the ratio difference accuracy index of the voltage transformer online monitoring device is:

[0080] in, It is the theoretical difference between the ratio error before adding the difference and the ratio error measured after adding the difference; The phase difference accuracy index of the voltage transformer online monitoring device is:

[0081] in, is the theoretical difference between the phase error before adding the difference and the phase error measured after adding the difference. If , , then the voltage transformer online monitoring device is evaluated to be normal, otherwise the voltage transformer online monitoring device is evaluated to be abnormal and needs to be repaired. The comma in the upper right corner of is the expression of the phase angle "minutes", Represents a phase angle of 2 minutes.

[0082] Different ratio differences and angle differences are added to the 12 voltage transformers. The comparison between the evaluation results of the voltage transformer online monitoring device and the difference addition results is shown in Table 1.

[0083] Table 1 Evaluation results and differential addition results of voltage transformer online monitoring device

[0084] The experimental results show that the voltage transformer online monitoring device can correctly identify the differential error, and the maximum deviation does not exceed 0.026%, -0.43′.

[0085] Through the above technical scheme, the present invention proposes a traceability platform for the accuracy of the evaluation of the online monitoring device of the voltage transformer, and proposes a phase micro-difference algorithm based on the combination of global and local features for the accuracy of phase difference addition in the portable traceability platform, thereby realizing the precise addition of phase.

[0086] Example 2 Based on Embodiment 1, Embodiment 2 of the present invention further provides an evaluation system for an online monitoring device of a voltage transformer, comprising: The first evaluation module is used to input voltage signals to each channel of the voltage transformer online monitoring device without adding ratio difference and phase difference, so as to complete error evaluation of each channel of the voltage transformer online monitoring device; The second evaluation module is used to select some channels of the voltage transformer online monitoring device in turn to perform differential addition when the input voltage signal remains unchanged, so as to complete the error evaluation of each channel of the voltage transformer online monitoring device; the differential addition refers to adding ratio difference and phase difference; The accuracy assessment module is used to calculate the ratio difference accuracy index and the phase difference accuracy index using the error assessment results before and after the difference addition. If the ratio difference accuracy index is less than or equal to the first preset value and the phase difference accuracy index is less than or equal to the second preset value, the voltage transformer online monitoring device is evaluated normally; otherwise, the voltage transformer online monitoring device is evaluated abnormally and requires maintenance.

[0087] Specifically, the second evaluation module selects some channels of the voltage transformer online monitoring device to perform analog-to-digital conversion on the input voltage signal before adding the difference.

[0088] Specifically, the method for adding phase difference to some channels of the voltage transformer online monitoring device in the second evaluation module is: S201, preprocessing the original signal to obtain a reference wave, and dividing the original signal into a plurality of sub-waves; S202, calculating similarity features between each wavelet and the reference wave, and calculating similarity between each wavelet and the reference wave according to the similarity features, wherein the similarity features include distance features, shape features, and data features; S203, selecting a wavelet whose similarity exceeds a set threshold as a candidate wavelet based on the similarity between each wavelet and the reference wave; selecting a band on the left side of the candidate wavelet as a first candidate extended wavelet, and selecting the first candidate extended wavelet according to the predicted maximum, minimum, and slope of the left extended wave of the original signal and the maximum, minimum, and slope of each first candidate extended wavelet, to obtain the left extended wave of the original signal; S204, selecting the waveband on the right side of the candidate wavelet as the second candidate extended wave, and selecting the second candidate extended wave according to the predicted maximum, minimum, slope of the right side extended wave of the original signal and the maximum, minimum, and slope of each second candidate extended wave, to obtain the right side extended wave of the original signal; S205, based on the left side extended wave of the original signal and the right side extended wave of the original signal, the original signal is extended to the left and right respectively, and the extended waveform is subjected to windowed Hilbert transform to realize the phase addition and difference of the corresponding channel of the voltage transformer online monitoring device.

[0089] Specifically, the S201 includes: Suppose the original signal is e(N), and the corresponding left endpoint is , past Draw a line parallel to the horizontal axis, and its intersection points with e(N) are , , is the total number of intersection points; As the starting point, the intercept length is The signal is used as the reference wave , when selected Contains a maximum point, a minimum point and a zero crossing point. is the number of sampling points of the reference wave, The reference wave The amplitude of the sampling points; As the starting point, the original signal is divided into K sub-waves , ,... , each sub-wave contains a maximum point, a minimum point and a zero-crossing point, and the number of sampling points of each sub-wave is recorded as .

[0090] Specifically, the process of S202 is: For Sub-wave , which is consistent with the reference wave The distance feature is

[0091] in, For the The number of sampling points of a wavelet, For the The sub-wave The amplitude of the sampling points, Indicates the calculation process from point (1,1) to point The cumulative regularization distance is calculated as follows:

[0092] In the formula, Indicates Sub-wave With Sub-wave The Euclidean distance between Before calculating the shape features between different waveforms, the waveform data of each sub-wave needs to be interpolated or sampled according to the reference wave to adjust the number of sampling points of the sub-wave to be consistent with that of the reference wave. The interpolation or sampling ratio of the wavelets is: ; Compute shape features:

[0093] in, is the sampling point number and =1,2,..., , , They are the reference wave, The sub-wave The amplitude of the sampling points, is an exponential function with base e; Calculate data features:

[0094] in, , They are the reference wave, The mean amplitude of the sub-waves at the sampling points; The similarity between each wavelet and the reference wave is calculated based on the similarity characteristics. The similarity calculation formula between a sub-wave and the reference wave is:

[0095] in, They are the set first weight, second weight and third weight respectively.

[0096] More specifically, the process of S203 is: Based on the similarity between each sub-wave and the reference wave, Select the corresponding wavelet as the candidate wavelet, where For the set threshold, select the band on the left side of the candidate sub-wave as the first candidate extension wave:

[0097] in, represents the total number of first candidate extension waves, is the u-th first candidate extension wave, , Indicates the Uth first candidate extension wave The amplitude of the sampling points; The horizontal and vertical coordinates of the maximum value of the left extension wave of the original signal are predicted ( ), the horizontal and vertical coordinates of the minimum value ( ), slope ; At the same time, the maximum, minimum, and slope of each first candidate extension wave are calculated, and the left extension wave of the original signal is selected from the first candidate extension wave based on the following formula:

[0098]

[0099] in, is the uth wave to be selected, Indicate point With point The Euclidean distance of are the ordinates of the maximum and minimum of the u-th first candidate extended wave, respectively. is the slope of the u-th first candidate extension wave, For the fourth and fourth and fifth weights, It is the left extension wave of the original signal.

[0100] More specifically, the prediction obtains the horizontal and vertical coordinates of the maximum value of the left extension wave of the original signal ( ), the horizontal and vertical coordinates of the minimum value ( ), slope The process is: Assume the original signal is (N), build a maximum value data set , Minimum Dataset , slope set :

[0101]

[0102]

[0103] in, is the number of maximum value data sets, minimum value data sets and slope concentrated data sets of the original signal, Represents the horizontal and vertical coordinate values ​​of the nth maximum point of the original signal from left to right, Represents the horizontal and vertical coordinate values ​​of the nth minimum point of the original signal from left to right, It represents the nth slope of the original signal from left to right. ; The LSTM model is used to predict the maximum point, minimum point and slope of the original signal, and based on the above data set , , The loss function is constructed based on the prediction results of the LSTM model, and the LSTM model is trained. When the loss function is the smallest, the training is stopped to obtain the trained LSTM model. The loss function is:

[0104] Where j represents the jth training sample, j=1,2,...J; Represents the number of model training samples. The jth training sample includes the horizontal and vertical coordinate values ​​of the jth maximum point of the original signal from left to right. , the horizontal and vertical coordinates of the jth minimum point of the original signal from left to right And the jth slope of the original signal from left to right ; ( ) represents the horizontal and vertical coordinates predicted values ​​of the jth maximum point of the original signal from left to right, ( ) represents the horizontal and vertical coordinates predicted values ​​of the jth minimum point of the original signal from left to right, Represents the predicted value of the jth slope of the original signal from left to right; Use the trained LSTM model to predict the horizontal and vertical coordinates of the maximum value of the left extension wave of the original signal ( ), the horizontal and vertical coordinates of the minimum value ( ), slope .

[0105] More specifically, the accuracy assessment module is further used to: Before adding the difference, the evaluation result of the voltage transformer online monitoring device on the i-th voltage transformer is: , represents the ratio error evaluation result of the i-th voltage transformer before adding the difference, represents the phase error evaluation result of the i-th voltage transformer before adding the difference; the evaluation result of the voltage transformer online monitoring device on the i-th voltage transformer after adding the difference is , represents the ratio error evaluation result of the i-th voltage transformer after adding the difference, represents the phase error evaluation result of the i-th voltage transformer after adding the difference, then the ratio difference accuracy index of the voltage transformer online monitoring device is:

[0106] in, It is the theoretical difference between the ratio error before adding the difference and the ratio error measured after adding the difference; The phase difference accuracy index of the voltage transformer online monitoring device is:

[0107] in, is the theoretical difference between the phase error before adding the difference and the phase error measured after adding the difference. If , , then the voltage transformer online monitoring device is evaluated to be normal, otherwise the voltage transformer online monitoring device is evaluated to be abnormal and needs to be repaired, among which, Represents a phase angle of 2 minutes.

[0108] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for evaluating an online monitoring device for a voltage transformer, characterized in that: include: S1. Without adding ratio difference and phase difference, a voltage signal is input to each channel of the voltage transformer online monitoring device to complete error evaluation of each channel of the voltage transformer online monitoring device; S2. When the input voltage signal remains unchanged, select some channels of the voltage transformer online monitoring device in turn to perform difference addition, so as to complete the error evaluation of each channel of the voltage transformer online monitoring device; The addition of difference refers to adding ratio difference and phase difference; S3. Calculate the ratio difference accuracy index and the phase difference accuracy index using the error evaluation results before and after the difference addition. If the ratio difference accuracy index is less than or equal to the first preset value and the phase difference accuracy index is less than or equal to the second preset value, the voltage transformer online monitoring device is evaluated normally. Otherwise, the voltage transformer online monitoring device is evaluated abnormally and requires maintenance.

2. The evaluation method of a voltage transformer online monitoring device according to claim 1, characterized in that: In the S2, some channels of the voltage transformer online monitoring device are selected to perform analog-to-digital conversion on the input voltage signal before adding the difference.

3. The evaluation method of a voltage transformer online monitoring device according to claim 1, characterized in that: The method for adding phase difference to some channels of the voltage transformer online monitoring device in S2 is: S201, preprocessing the original signal to obtain a reference wave, and dividing the original signal into a plurality of sub-waves; S202, calculating similarity features between each wavelet and the reference wave, and calculating similarity between each wavelet and the reference wave according to the similarity features, wherein the similarity features include distance features, shape features, and data features; S203, selecting a wavelet whose similarity exceeds a set threshold as a candidate wavelet based on the similarity between each wavelet and the reference wave; selecting a band on the left side of the candidate wavelet as a first candidate extended wavelet, and selecting the first candidate extended wavelet according to the predicted maximum, minimum, and slope of the left extended wave of the original signal and the maximum, minimum, and slope of each first candidate extended wavelet, to obtain the left extended wave of the original signal; S204, selecting the waveband on the right side of the candidate wavelet as the second candidate extended wave, and selecting the second candidate extended wave according to the predicted maximum, minimum, slope of the right side extended wave of the original signal and the maximum, minimum, and slope of each second candidate extended wave, to obtain the right side extended wave of the original signal; S205, based on the left side extended wave of the original signal and the right side extended wave of the original signal, the original signal is extended to the left and right respectively, and the extended waveform is subjected to windowed Hilbert transform to realize the phase addition and difference of the corresponding channel of the voltage transformer online monitoring device.

4. The evaluation method for an online monitoring device for a voltage transformer according to claim 3 is characterized in that: The S201 includes: Suppose the original signal is e(N), and the corresponding left endpoint is , past Draw a line parallel to the horizontal axis, and its intersection points with e(N) are , , is the total number of intersection points; As the starting point, the intercept length is The signal is used as the reference wave , when selected Contains a maximum point, a minimum point and a zero crossing point. is the number of sampling points of the reference wave, The reference wave The amplitude of the sampling points; As the starting point, the original signal is divided into K sub-waves , ,... , each sub-wave contains a maximum point, a minimum point and a zero-crossing point, and the number of sampling points of each sub-wave is recorded as .

5. The evaluation method for an online monitoring device for a voltage transformer according to claim 4, characterized in that: The process of S202 is as follows: For Sub-wave , which is consistent with the reference wave The distance feature is in, For the The number of sampling points of a wavelet, For the The sub-wave The amplitude of the sampling points, Indicates the calculation process from point (1,1) to point The cumulative regularization distance is calculated as follows: In the formula, Indicates Sub-wave With Sub-wave The Euclidean distance between Before calculating the shape features between different waveforms, the waveform data of each sub-wave needs to be interpolated or sampled according to the reference wave to adjust the number of sampling points of the sub-wave to be consistent with that of the reference wave. The interpolation or sampling ratio of the wavelets is: ; Compute shape features: in, is the sampling point number and =1,2,..., , , They are the reference wave, The sub-wave The amplitude of the sampling points, is an exponential function with base e; Calculate data features: in, , They are the reference wave, The mean amplitude of the sub-waves at the sampling points; The similarity between each wavelet and the reference wave is calculated based on the similarity characteristics. The similarity calculation formula between a sub-wave and the reference wave is: in, They are the set first weight, second weight and third weight respectively.

6. The evaluation method of a voltage transformer online monitoring device according to claim 5, characterized in that: The process of S203 is as follows: Based on the similarity between each sub-wave and the reference wave, Select the corresponding wavelet as the candidate wavelet, where For the set threshold, select the band on the left side of the candidate sub-wave as the first candidate extension wave: in, represents the total number of first candidate extension waves, is the u-th first candidate extension wave, , Indicates the Uth first candidate extension wave The amplitude of the sampling points; The horizontal and vertical coordinates of the maximum value of the left extension wave of the original signal are predicted ( ), the horizontal and vertical coordinates of the minimum value ( ), slope ; At the same time, the maximum, minimum, and slope of each first candidate extension wave are calculated, and the left extension wave of the original signal is selected from the first candidate extension wave based on the following formula: in, is the uth wave to be selected, Indicate point With point The Euclidean distance of are the ordinates of the maximum and minimum of the u-th first candidate extended wave, respectively. is the slope of the u-th first candidate extension wave, For the fourth and fourth and fifth weights, It is the left extension wave of the original signal.

7. The evaluation method of a voltage transformer online monitoring device according to claim 6, characterized in that: The prediction obtains the horizontal and vertical coordinates of the maximum value of the left extension wave of the original signal ( ), the horizontal and vertical coordinates of the minimum value ( ), slope The process is: Assume the original signal is (N), build a maximum value data set , Minimum Dataset , slope set : in, is the number of maximum value data sets, minimum value data sets and slope concentrated data sets of the original signal, Represents the horizontal and vertical coordinate values ​​of the nth maximum point of the original signal from left to right, Represents the horizontal and vertical coordinate values ​​of the nth minimum point of the original signal from left to right, It represents the nth slope of the original signal from left to right. ; The LSTM model is used to predict the maximum point, minimum point and slope of the original signal, and based on the above data set , , The loss function is constructed based on the prediction results of the LSTM model, and the LSTM model is trained. When the loss function is the smallest, the training is stopped to obtain the trained LSTM model. The loss function is: Where j represents the jth training sample, j=1,2,...J; Represents the number of model training samples. The jth training sample includes the horizontal and vertical coordinate values ​​of the jth maximum point of the original signal from left to right. , the horizontal and vertical coordinates of the jth minimum point of the original signal from left to right And the jth slope of the original signal from left to right ; ( ) represents the horizontal and vertical coordinates predicted values ​​of the jth maximum point of the original signal from left to right, ( ) represents the horizontal and vertical coordinates predicted values ​​of the jth minimum point of the original signal from left to right, Represents the predicted value of the jth slope of the original signal from left to right; Use the trained LSTM model to predict the horizontal and vertical coordinates of the maximum value of the left extension wave of the original signal ( ), the horizontal and vertical coordinates of the minimum value ( ), slope .

8. The evaluation method for an online monitoring device for a voltage transformer according to claim 6, characterized in that: The process of S3 is as follows: Before adding the difference, the evaluation result of the voltage transformer online monitoring device on the i-th voltage transformer is: , represents the ratio error evaluation result of the i-th voltage transformer before adding the difference, represents the phase error evaluation result of the i-th voltage transformer before adding the difference; the evaluation result of the voltage transformer online monitoring device on the i-th voltage transformer after adding the difference is , represents the ratio error evaluation result of the i-th voltage transformer after adding the difference, represents the phase error evaluation result of the i-th voltage transformer after adding the difference, then the ratio difference accuracy index of the voltage transformer online monitoring device is: in, It is the theoretical difference between the ratio error before adding the difference and the ratio error measured after adding the difference; The phase difference accuracy index of the voltage transformer online monitoring device is: in, is the theoretical difference between the phase error before adding the difference and the phase error measured after adding the difference. If , , then the voltage transformer online monitoring device is evaluated to be normal, otherwise the voltage transformer online monitoring device is evaluated to be abnormal and needs to be repaired, among which, Represents a phase angle of 2 minutes.

9. An evaluation system for an online monitoring device of a voltage transformer, characterized in that: include: The first evaluation module is used to input voltage signals to each channel of the voltage transformer online monitoring device without adding ratio difference and phase difference, so as to complete error evaluation of each channel of the voltage transformer online monitoring device; The second evaluation module is used to select some channels of the voltage transformer online monitoring device for difference addition when the input voltage signal remains unchanged, so as to complete the error evaluation of each channel of the voltage transformer online monitoring device; The addition of difference refers to adding ratio difference and phase difference; The accuracy assessment module is used to calculate the ratio difference accuracy index and the phase difference accuracy index using the error assessment results before and after the difference addition. If the ratio difference accuracy index is less than or equal to the first preset value and the phase difference accuracy index is less than or equal to the second preset value, the voltage transformer online monitoring device is evaluated normally; otherwise, the voltage transformer online monitoring device is evaluated abnormally and requires maintenance.

10. An evaluation system for an online monitoring device for a voltage transformer according to claim 9, characterized in that: The second evaluation module selects some channels of the voltage transformer online monitoring device to perform analog-to-digital conversion on the input voltage signal before adding the difference.

Images