Denoising method and device for partial discharge detection and medium

Abstract

The embodiment of the invention discloses a partial discharge detection denoising method, partial discharge detection denoising equipment and a medium, belongs to the technical field of partial discharge detection, and solves the problem that an existing partial discharge detection denoising method damages the integrity and accuracy of partial discharge signals. Drawing a time sequence energy instantaneous change rate curve; windowing the time sequence energy instantaneous change rate curve, comparing the time sequence energy instantaneous change rate corresponding to the data points in a window with a preset filtering threshold value, and performing primary denoising on the high-frequency current signal to be processed based on a comparison result; in the time sequence energy instantaneous change rate curve after primary denoising, determining a data point with the time sequence energy instantaneous change rate greater than zero so as to extract a pulse waveform from the to-be-processed high-frequency current signal and construct a pulse array; and determining a spectral bandwidth corresponding to the pulse array at a preset amplitude ratio threshold, and performing secondary denoising on the to-be-processed high-frequency current signal based on the spectral bandwidth to obtain a partial discharge signal.

Description

Technical Field

[0001] The present application relates to the technical field of partial discharge detection, and in particular to a denoising method, device and medium for partial discharge detection. Background Art

[0002] During operation, converter transformers are subjected to complex stresses such as electricity, heat and force. Equipment insulation is prone to defects such as bubbles, cracks, and burrs, leading to partial discharge. Partial discharge will aggravate the degradation of insulation materials, which may cause insulation failure and cause serious safety accidents. By detecting partial discharge, it is helpful to timely discover potential defects in converter transformers, avoid the occurrence of insulation failure, and effectively ensure the safety of equipment and the stable operation of the power grid.

[0003] The high-frequency current method is widely used in partial discharge detection at the converter station site, and the high-frequency current signal is collected through a high-frequency current transformer. However, due to the serious electromagnetic interference at the converter station site, the partial discharge high-frequency current signal collected at the site may be submerged, seriously affecting the accuracy of partial discharge detection. Therefore, in order to accurately detect partial discharge, it is first necessary to eliminate the influence of noise interference.

[0004] The noise interference at the converter station site can be mainly divided into three categories: white noise, periodic narrowband interference and random pulse interference. White noise generally comes from the environment or the equipment itself, and is usually considered to satisfy the Gaussian distribution. Periodic narrowband interference is mainly generated by carrier communication, higher harmonics and radio communication, and it has a narrowband spectrum centered on the main frequency. Random pulse interference is generally a broadband signal caused by the action of thyristors and other switching devices. Once the frequency of these noise interferences overlaps with the frequency of the partial discharge signal, it is difficult to effectively remove the interference. For low-frequency noise interference, the frequency domain filtering method can certainly remove it, but it will also filter out the low-frequency component of the partial discharge signal, resulting in the omission of discharge information, causing the partial discharge signal to attenuate, and destroying the integrity and accuracy of the partial discharge signal. Summary of the invention

[0005] The embodiments of the present application provide a denoising method, device and medium for partial discharge detection, which are used to solve the following technical problems: the existing denoising method for partial discharge detection easily leads to omission of discharge information, causing partial discharge signal attenuation, thereby destroying the integrity and accuracy of the partial discharge signal.

[0006] The present application embodiment adopts the following technical solutions:

[0007] The embodiment of the present application provides a denoising method for partial discharge detection. The method includes determining the instantaneous change rate of the time series energy of the high-frequency current signal to be processed, and drawing a curve of the instantaneous change rate of the time series energy; performing windowing processing on the instantaneous change rate of the time series energy, comparing the instantaneous change rate of the time series energy corresponding to the data point in the window with the preset filtering threshold, and performing a primary denoising on the high-frequency current signal to be processed based on the comparison result; in the instantaneous change rate of the time series energy curve after the primary denoising, determining the data point whose instantaneous change rate of the time series energy is greater than zero, and extracting the pulse waveform from the high-frequency current signal to be processed based on the data point whose instantaneous change rate of the time series energy is greater than zero, and constructing a pulse array; determining the spectral bandwidth corresponding to each pulse waveform in the pulse array at the preset amplitude ratio threshold, and performing secondary denoising on the high-frequency current signal to be processed based on the spectral bandwidth to obtain a partial discharge signal.

[0008] The embodiment of the present application uses the instantaneous change rate of the time series energy of the signal as the identification basis, and can effectively distinguish white noise, periodic narrowband interference and pulse waveforms, so as to adaptively extract the pulse waveform. Secondly, according to the spectrum difference between the local discharge signal and the interference pulse, the embodiment of the present application proposes the spectrum bandwidth corresponding to each pulse waveform in the pulse array at the preset amplitude ratio threshold as an indicator to distinguish the local discharge signal from the interference signal, which can realize the adaptive identification of the local discharge signal with higher identification accuracy. The embodiment of the present application can denoise different types of noise interference existing in the high-frequency current method detection, and will not attenuate the local discharge signal, and the calculation amount is small, which can meet the needs of rapid detection of local discharge on-site in the converter station.

[0009] In one implementation of the present application, the instantaneous change rate of the time series energy of the high-frequency current signal to be processed is determined, specifically including: based on a preset time series energy accumulation rate function, determining the time series energy accumulation rate corresponding to each data point in the high-frequency signal to be processed; based on the difference between the time series energy accumulation rate corresponding to the k+1th data point and the time series energy accumulation rate corresponding to the kth data point, obtaining the instantaneous change rate of the time series energy.

[0010] In one implementation of the present application, the instantaneous change rate of the time series energy corresponding to the data points in the window is compared with the preset filtering threshold, and the high-frequency current signal to be processed is denoised once based on the comparison result, specifically including: within the window, determining the number of data points whose instantaneous change rate of the time series energy is greater than the preset filtering threshold; when the number of data points is greater than the data volume proportion threshold corresponding to the window, determining that there is a pulse waveform in the window, and retaining the curve corresponding to the window; otherwise, setting the amplitude of the curve corresponding to the window to zero to perform a denoising on the high-frequency current signal to be processed; wherein, the denoising at least includes removing white noise and periodic narrowband interference in the high-frequency current signal to be processed.

[0011] In one implementation of the present application, before comparing the instantaneous change rate of the time series energy corresponding to the data points in the window with the preset filtering threshold, the method also includes: sorting the instantaneous change rate of the time series energy corresponding to all the data points in the window from small to large; among the sorted data points, determining the instantaneous change rate of the time series energy corresponding to the data points at the preset proportion position, and using the instantaneous change rate of the time series energy corresponding to the data points at the preset proportion position as the preset filtering threshold.

[0012] In one implementation of the present application, in the instantaneous change rate curve of the time series energy after a denoising, the data points whose instantaneous change rate of the time series energy is greater than zero are determined, and based on the data points whose instantaneous change rate of the time series energy is greater than zero, the pulse waveform is extracted from the high-frequency current signal to be processed, and a pulse array is constructed, which specifically includes: determining the data points whose instantaneous change rate of the time series energy is greater than zero; based on a preset judgment function, determining the first time period corresponding to the data points whose instantaneous change rate of the time series energy is zero, and setting the high-frequency current signal to be processed corresponding to the first time period to zero; based on the preset judgment function, determining the second time period corresponding to the data points whose instantaneous change rate of the time series energy is greater than zero, and retaining the high-frequency current signal to be processed corresponding to the second time period to obtain a pulse array.

[0013] In one implementation of the present application, the preset judgment function is:

[0014]

[0015] Among them, P(i) represents the amplitude corresponding to the i-th data point of the pulse array, x(i) represents the amplitude corresponding to the i-th data point of the high-frequency current signal to be processed; ΔE(i) represents the instantaneous change rate of the time series energy corresponding to the i-th data point.

[0016] In one implementation of the present application, the spectral bandwidth corresponding to each pulse waveform in the pulse array at a preset amplitude ratio threshold is determined, and the high-frequency current signal to be processed is subjected to secondary denoising based on the spectral bandwidth to obtain a local discharge signal, specifically including: determining the spectrum corresponding to each pulse waveform in the pulse array, and determining the maximum spectrum amplitude; determining the reference spectrum amplitude corresponding to the preset amplitude ratio threshold of the maximum spectrum amplitude; and determining the first frequency and the second frequency corresponding to the reference spectrum amplitude; based on the spectral bandwidth between the first frequency and the second frequency, the high-frequency current signal to be processed is subjected to secondary denoising to obtain a local discharge signal.

[0017] In one implementation of the present application, based on the spectrum bandwidth between the first frequency and the second frequency, the high-frequency current signal to be processed is subjected to secondary denoising to obtain a local discharge signal, specifically including: when the spectrum bandwidth between the first frequency and the second frequency is greater than a preset spectrum bandwidth threshold, the pulse waveform is determined to be a local discharge signal and is retained; when the spectrum bandwidth between the first frequency and the second frequency is not greater than the preset spectrum bandwidth threshold, the pulse waveform is determined to be an interference pulse, and the interference pulse is subjected to amplitude zeroing processing.

[0018] The embodiment of the present application provides a denoising device for partial discharge detection, comprising: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can: determine the instantaneous change rate of the time series energy of the high-frequency current signal to be processed, and draw a curve of the instantaneous change rate of the time series energy; perform windowing processing on the instantaneous change rate of the time series energy curve, compare the instantaneous change rate of the time series energy corresponding to the data point in the window with a preset filtering threshold, and perform a primary denoising on the high-frequency current signal to be processed based on the comparison result; in the instantaneous change rate of the time series energy curve after the primary denoising, determine the data points whose instantaneous change rate of the time series energy is greater than zero, and extract pulse waveforms from the high-frequency current signal to be processed based on the data points whose instantaneous change rate of the time series energy is greater than zero, and construct a pulse array; determine the spectrum bandwidth corresponding to each pulse waveform in the pulse array at the preset amplitude proportion threshold, and perform secondary denoising on the high-frequency current signal to be processed based on the spectrum bandwidth to obtain a partial discharge signal.

[0019] A non-volatile computer storage medium provided in an embodiment of the present application stores computer executable instructions, wherein the computer executable instructions are configured to: determine the instantaneous change rate of the time series energy of a high-frequency current signal to be processed, and draw a curve of the instantaneous change rate of the time series energy; perform windowing processing on the instantaneous change rate of the time series energy, compare the instantaneous change rate of the time series energy corresponding to the data point in the window with a preset filtering threshold, and perform a primary denoising on the high-frequency current signal to be processed based on the comparison result; determine the data points whose instantaneous change rate of the time series energy is greater than zero in the instantaneous change rate of the time series energy curve after the primary denoising, and extract pulse waveforms from the high-frequency current signal to be processed based on the data points whose instantaneous change rate of the time series energy is greater than zero, and construct a pulse array; determine the spectral bandwidth corresponding to each pulse waveform in the pulse array at a preset amplitude ratio threshold, and perform secondary denoising on the high-frequency current signal to be processed based on the spectral bandwidth to obtain a partial discharge signal.

[0020] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects: the embodiments of the present application use the instantaneous change rate of the time series energy of the signal as the identification basis, and can effectively distinguish white noise, periodic narrowband interference and pulse waveforms, so as to adaptively extract the pulse waveform. Secondly, according to the spectrum difference between the local discharge signal and the interference pulse, the embodiments of the present application propose to use the spectrum bandwidth corresponding to each pulse waveform in the pulse array at the preset amplitude ratio threshold as an indicator to distinguish the local discharge signal from the interference signal, so as to realize the adaptive identification of the local discharge signal with higher identification accuracy. The embodiments of the present application can denoise different types of noise interference existing in the high-frequency current method detection, and will not attenuate the local discharge signal, and the calculation amount is small, which can meet the needs of rapid detection of local discharge at the converter station site. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. In the drawings:

[0022] Figure 1 A flow chart of a denoising method for partial discharge detection provided in an embodiment of the present application;

[0023] Figure 2 A schematic diagram of a high-frequency current signal to be processed provided in an embodiment of the present application;

[0024] Figure 3 A schematic diagram of the instantaneous rate of change of the time series energy of a signal provided in an embodiment of the present application;

[0025] Figure 4 A schematic diagram of a pulse array provided in an embodiment of the present application;

[0026] Figure 5 A schematic diagram of the -20dB spectrum bandwidth of each pulse waveform provided in an embodiment of the present application;

[0027] Figure 6 A schematic diagram of an overall denoising result provided in an embodiment of the present application;

[0028] Figure 7 A schematic diagram of a local amplification denoising result provided in an embodiment of the present application;

[0029] Figure 8 A schematic diagram of the structure of a denoising device for partial discharge detection provided in an embodiment of the present application. DETAILED DESCRIPTION

[0030] Embodiments of the present application provide a denoising method, device and medium for partial discharge detection.

[0031] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0032] At present, the main denoising methods used in partial discharge detection include wavelet packet denoising, fast Fourier transform threshold method, variational mode decomposition and singular value decomposition. The main principle of these methods is to decompose the noisy signal according to different standards, remove the noise components and reconstruct the signal to achieve denoising. However, these methods can only remove white noise or periodic narrowband interference in a single way, which is difficult to meet the denoising requirements of complex noise interference on site. At the same time, these methods have limitations in practical applications due to the need to manually set the number of decomposition layers, the attenuation of partial discharge signals and the large amount of calculation. In summary, denoising algorithms that can effectively remove multiple types of noise interference and do not attenuate partial discharge signals still need further research.

[0033] The embodiment of the present application proposes a denoising method for partial discharge detection, which can effectively remove noise interference at the converter station site without attenuating the partial discharge signal. First, the time series energy accumulation rate of the signal is calculated, and then the instantaneous change rate of the time series energy is obtained; then, the corresponding waveform is extracted from the original signal according to the instantaneous change rate to obtain a pulse array; finally, each pulse waveform is analyzed to calculate the bandwidth corresponding to the spectrum amplitude attenuating 20dB from the maximum value. According to the difference in the bandwidth of the partial discharge pulse and the interference pulse, the partial discharge signal is identified and the pulse interference is eliminated, thereby realizing the denoising of partial discharge detection.

[0034] The technical solution proposed in the embodiment of the present invention is described in detail below with reference to the accompanying drawings.

[0035] Figure 1 A flow chart of a denoising method for partial discharge detection provided in an embodiment of the present application, such as Figure 1 As shown, the denoising method for partial discharge detection includes the following steps:

[0036] S101, determining the instantaneous change rate of the time series energy of the high-frequency current signal to be processed, and drawing a curve of the instantaneous change rate of the time series energy.

[0037] In one embodiment of the present application, based on a preset time series energy accumulation rate function, the time series energy accumulation rate corresponding to each data point in the high-frequency signal to be processed is determined. Based on the difference between the time series energy accumulation rate corresponding to the k+1th data point and the time series energy accumulation rate corresponding to the kth data point, the instantaneous change rate of the time series energy is obtained.

[0038] Specifically, the time series energy accumulation rate is used to describe the distribution of signal energy over time. The time series energy accumulation rate of the signal at the kth data point is defined as follows:

[0039]

[0040] Where N is the number of sampling points of the signal, and x(i) represents the amplitude corresponding to the i-th data point of the signal. The denominator represents the total energy of the signal. Normalization based on the total energy of the signal can eliminate the influence of the signal amplitude.

[0041] The instantaneous rate of change of the time series energy ΔE reflects the instantaneous growth of the signal time series energy and is calculated according to the following formula:

[0042] ΔE=E(k+1)-E(k) (2)

[0043] Among them, E(k) represents the time series energy accumulation rate corresponding to the kth data point of the signal; E(k+1) represents the time series energy accumulation rate corresponding to the k+1th data point of the signal; ΔE is the instantaneous change rate of the time series energy.

[0044] S102, performing windowing processing on the instantaneous change rate curve of the time series energy, comparing the instantaneous change rate of the time series energy corresponding to the data points in the window with a preset filtering threshold, and performing a denoising on the high-frequency current signal to be processed based on the comparison result.

[0045] In one embodiment of the present application, the instantaneous change rates of the time series energy corresponding to all data points in the window are sorted from small to large. Among the sorted data points, the instantaneous change rate of the time series energy corresponding to the data point at the preset proportion position is determined, and the instantaneous change rate of the time series energy corresponding to the data point at the preset proportion position is used as the preset filtering threshold.

[0046] Specifically, for white noise and periodic narrowband interference, the instantaneous change rate of the signal's time series energy is very small. In order to remove this part of noise interference, the embodiment of the present application sets the filtering threshold as follows:

[0047] ΔE T =ΔE 80％ (3)

[0048] Where, ΔE T is the preset filtering threshold; ΔE 80％It means that the instantaneous change rates of the time series energy of all data points of the signal are sorted in order from small to large, and the instantaneous change rates of the time series energy corresponding to the data points that are just greater than 80% are found.

[0049] It should be noted that the embodiment of the present application preferably uses the instantaneous change rate of the time series energy corresponding to data points greater than 80% as the preset filtering threshold, which can be adjusted according to actual conditions in the application, and the embodiment of the present application does not limit this.

[0050] In one embodiment of the present application, within the window, the number of data points whose instantaneous rate of change of time series energy is greater than a preset filtering threshold is determined. When the number of data points is greater than the data volume percentage threshold corresponding to the window, it is determined that there is a pulse waveform in the window, and the curve corresponding to the window is retained. Otherwise, the amplitude of the curve corresponding to the window is set to zero to perform a denoising on the high-frequency current signal to be processed. Among them, the denoising at least includes removing white noise and periodic narrowband interference in the high-frequency current signal to be processed.

[0051] Specifically, the instantaneous change rate of the time series energy is windowed, and the window size can be 150 data points, or it can be adjusted according to the actual application. If the instantaneous change rate of the time series energy corresponding to more than 20% of the data points in the window is greater than the preset filtering threshold ΔE T , it is considered that there is a pulse waveform in the window and it is retained, otherwise the curve amplitude in the window is set to zero.

[0052] Among them, 20% is the preferred data volume proportion threshold set in the embodiment of the present application. It can be adjusted according to actual conditions in the application, and the embodiment of the present application does not limit this.

[0053] S103. In the instantaneous change rate curve of the time series energy after one denoising, determine the data points where the instantaneous change rate of the time series energy is greater than zero, and extract the pulse waveform from the high-frequency current signal to be processed based on the data points where the instantaneous change rate of the time series energy is greater than zero to construct a pulse array.

[0054] In one embodiment of the present application, a data point whose instantaneous change rate of the time series energy is greater than zero is determined; based on a preset judgment function, a first time period corresponding to the data point whose instantaneous change rate of the time series energy is zero is determined, and the high-frequency current signal to be processed corresponding to the first time period is set to zero; based on the preset judgment function, a second time period corresponding to the data point whose instantaneous change rate of the time series energy is greater than zero is determined, and the high-frequency current signal to be processed corresponding to the second time period is retained to obtain a pulse array.

[0055] Specifically, after removing white noise and periodic narrowband interference, the signal is traversed, and the high-frequency current signal to be processed in the time period corresponding to the data point with zero instantaneous change rate of time series energy is directly set to zero; the high-frequency current signal to be processed in the time period corresponding to the data point with instantaneous change rate of time series energy greater than zero is retained, and finally a pulse array is obtained.

[0056] Specifically, according to preset conditions, such as the absolute value of the rate of change is less than a certain threshold, the data points with a rate of change close to zero are divided into one or more time periods, and for each identified first time period, the corresponding value of the high-frequency current signal to be processed is set to zero. According to the same preset judgment function, the time period with a significant increase in energy, that is, the second time period, is identified, and for each second time period, the corresponding high-frequency current signal to be processed is kept unchanged. After the above steps, the original high-frequency current signal to be processed is processed into a new signal, in which only the part with a significant increase in energy is retained, and the part with stable energy or little change is set to zero, thereby obtaining a pulse array.

[0057] In one embodiment of the present application, the preset judgment function is:

[0058]

[0059] Among them, P(i) represents the amplitude corresponding to the i-th data point of the pulse array, x(i) represents the amplitude corresponding to the i-th data point of the high-frequency current signal to be processed; ΔE(i) represents the instantaneous change rate of the time series energy corresponding to the i-th data point.

[0060] S104, determining the spectrum bandwidth corresponding to each pulse waveform in the pulse array at a preset amplitude ratio threshold, and performing secondary denoising on the high-frequency current signal to be processed based on the spectrum bandwidth to obtain a partial discharge signal.

[0061] In one embodiment of the present application, the spectrum corresponding to each pulse waveform in the pulse array is determined, and the maximum spectrum amplitude is determined. The reference spectrum amplitude corresponding to the preset amplitude ratio threshold of the maximum spectrum amplitude is determined; and the first frequency and the second frequency corresponding to the reference spectrum amplitude are determined. Based on the spectrum bandwidth between the first frequency and the second frequency, the high-frequency current signal to be processed is subjected to secondary denoising to obtain a partial discharge signal.

[0062] Specifically, when the spectrum bandwidth between the first frequency and the second frequency is greater than the preset spectrum bandwidth threshold, the pulse waveform is determined to be a partial discharge signal and is retained. When the spectrum bandwidth between the first frequency and the second frequency is not greater than the preset spectrum bandwidth threshold, the pulse waveform is determined to be an interference pulse and the interference pulse is subjected to amplitude zeroing processing.

[0063] Furthermore, the spectrum X(f) of each pulse waveform in the pulse array is calculated to obtain the maximum spectrum amplitude X(f) max When the spectrum amplitude is attenuated by 10 times (-20dB) from the maximum value, the corresponding bandwidth is the -20dB spectrum bandwidth, which is calculated according to the following formula:

[0064]

[0065] Among them, B W represents the -20dB spectrum bandwidth, f1 and f2 represent the frequency corresponding to one-tenth of the maximum spectrum amplitude. The -20dB spectrum bandwidth of each pulse waveform is calculated. If the bandwidth is greater than 10MHz, it is a partial discharge signal and is retained. If the bandwidth is less than 10MHz, it is an interference pulse and its amplitude is set to zero, thus completing the adaptive identification of partial discharge signals.

[0066] The following is a specific experimental verification:

[0067] (1) Figure 2 A schematic diagram of a high-frequency current signal to be processed is provided in an embodiment of the present application, which is a high-frequency current signal collected at the grounding point of the core of a converter transformer in a converter station.

[0068] (2) Figure 3 A schematic diagram of the instantaneous rate of change of the time series energy of a signal provided in an embodiment of the present application is shown in FIG. Figure 3 As shown, the horizontal axis is the signal time, and the vertical axis is the instantaneous change rate of the time series energy corresponding to each data point. Figure 3 The acquisition process is as follows: first determine Figure 2 The time series energy accumulation rate corresponding to each data point in the signal is calculated based on the time series energy accumulation rate.

[0069] (3) Figure 4 A pulse array schematic diagram is provided in an embodiment of the present application. The acquisition process of the pulse array schematic diagram is as follows: Figure 3 Windowing is performed, and within the window, the number of data points whose instantaneous rate of change of time series energy is greater than the preset filtering threshold is determined to remove white noise and periodic narrowband interference from the high-frequency current signal to be processed. Secondly, according to the preset judgment function, the pulse waveform of the corresponding time period is extracted from the high-frequency current signal to be processed to construct a pulse array.

[0070] (4) Figure 5 A schematic diagram of the -20dB spectrum bandwidth of each pulse waveform provided in an embodiment of the present application, such as Figure 5 As shown, the horizontal axis is the signal time, and the vertical axis is the -20dB spectrum bandwidth corresponding to each data point. Figure 6A schematic diagram of an overall denoising result provided in an embodiment of the present application is shown in FIG. Figure 7 A schematic diagram of a local amplification denoising result provided in an embodiment of the present application. Figure 6 and Figure 7 , is Figure 5 The result diagram obtained by denoising the pulse waveform with a bandwidth greater than 10 MHz is retained, thereby obtaining the partial discharge signal.

[0071] According to the waveform characteristics of white noise and periodic narrowband interference, the embodiment of the present application proposes to use the instantaneous change rate of the time series energy of the signal as the identification basis, which can effectively distinguish white noise, periodic narrowband interference and pulse waveform, so as to extract the pulse waveform adaptively. Secondly, according to the spectrum difference between the partial discharge signal and the interference pulse, the embodiment of the present application proposes to use the -20dB spectrum bandwidth of the pulse waveform as an indicator to distinguish the partial discharge signal from the interference signal, which can realize the adaptive identification of the partial discharge signal with high recognition accuracy. In addition, the method proposed in the embodiment of the present application adopts a time-frequency joint method for denoising, which has a good denoising effect for different types of noise interference existing in the high-frequency current method detection, and will not cause the partial discharge signal to attenuate. The embodiment of the present application has a small amount of calculation and can meet the needs of rapid detection of partial discharge on-site in the converter station.

[0072] Figure 8 A schematic diagram of the structure of a denoising device for partial discharge detection provided in an embodiment of the present application. Figure 8 As shown, a denoising device for partial discharge detection includes: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can: determine the instantaneous change rate of the time series energy of the high-frequency current signal to be processed, and draw a curve of the instantaneous change rate of the time series energy; perform windowing processing on the instantaneous change rate of the time series energy, compare the instantaneous change rate of the time series energy corresponding to the data point in the window with the preset filtering threshold, and perform a primary denoising on the high-frequency current signal to be processed based on the comparison result; in the instantaneous change rate of the time series energy curve after the primary denoising, determine the data points whose instantaneous change rate of the time series energy is greater than zero, and extract the pulse waveform from the high-frequency current signal to be processed based on the data points whose instantaneous change rate of the time series energy is greater than zero, and construct a pulse array; determine the spectrum bandwidth corresponding to each pulse waveform in the pulse array at the preset amplitude proportion threshold, and perform secondary denoising on the high-frequency current signal to be processed based on the spectrum bandwidth to obtain a partial discharge signal.

[0073] A non-volatile computer storage medium provided in an embodiment of the present application stores computer executable instructions, wherein the computer executable instructions are configured to: determine the instantaneous change rate of the time series energy of a high-frequency current signal to be processed, and draw a curve of the instantaneous change rate of the time series energy; perform windowing processing on the instantaneous change rate of the time series energy, compare the instantaneous change rate of the time series energy corresponding to the data point in the window with a preset filtering threshold, and perform a primary denoising on the high-frequency current signal to be processed based on the comparison result; determine the data points whose instantaneous change rate of the time series energy is greater than zero in the instantaneous change rate of the time series energy curve after the primary denoising, and extract pulse waveforms from the high-frequency current signal to be processed based on the data points whose instantaneous change rate of the time series energy is greater than zero, and construct a pulse array; determine the spectral bandwidth corresponding to each pulse waveform in the pulse array at a preset amplitude ratio threshold, and perform secondary denoising on the high-frequency current signal to be processed based on the spectral bandwidth to obtain a partial discharge signal.

[0074] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0075] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the embodiments of the present application may have various modifications and changes. 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 various embodiments of the present application.

Claims

1. A denoising method for partial discharge detection, characterized in that: The method comprises: Determine the instantaneous change rate of the time series energy of the high-frequency current signal to be processed, and draw a curve of the instantaneous change rate of the time series energy; Performing windowing processing on the instantaneous change rate curve of the time series energy, comparing the instantaneous change rate of the time series energy corresponding to the data point in the window with the preset filtering threshold, and performing a denoising on the high-frequency current signal to be processed based on the comparison result; In the instantaneous change rate curve of the time series energy after the first denoising, determine the data points where the instantaneous change rate of the time series energy is greater than zero, so as to extract the pulse waveform from the high-frequency current signal to be processed based on the data points where the instantaneous change rate of the time series energy is greater than zero, and construct a pulse array; The spectrum bandwidth corresponding to each pulse waveform in the pulse array at a preset amplitude proportion threshold is determined, and the high-frequency current signal to be processed is secondary denoised based on the spectrum bandwidth to obtain a partial discharge signal.

2. A denoising method for partial discharge detection according to claim 1, characterized in that: The step of determining the instantaneous change rate of the time series energy of the high-frequency current signal to be processed specifically includes: Based on a preset time series energy accumulation rate function, determining the time series energy accumulation rate corresponding to each data point in the high-frequency signal to be processed; The instantaneous change rate of the time series energy is obtained based on the difference between the time series energy accumulation rate corresponding to the k+1th data point and the time series energy accumulation rate corresponding to the kth data point.

3. A denoising method for partial discharge detection according to claim 1, characterized in that: The instantaneous change rate of the time series energy corresponding to the data points in the window is compared with the preset filtering threshold, and the high-frequency current signal to be processed is de-noised based on the comparison result, specifically including: In the window, determining the number of data points whose instantaneous rate of change of time series energy is greater than the preset filtering threshold; When the number of the data points is greater than the data volume percentage threshold corresponding to the window, it is determined that a pulse waveform exists in the window, and the curve corresponding to the window is retained; Otherwise, the curve amplitude corresponding to the window is set to zero to perform a denoising operation on the high-frequency current signal to be processed; wherein the denoising operation at least includes removing white noise and periodic narrowband interference in the high-frequency current signal to be processed.

4. A denoising method for partial discharge detection according to claim 1, characterized in that: Before comparing the instantaneous change rate of the time series energy corresponding to the data points in the window with the preset filtering threshold, the method further includes: Sort the instantaneous change rates of the time series energy corresponding to all data points in the window from small to large; Among the sorted data points, the instantaneous change rate of the time series energy corresponding to the data points at the preset proportion position is determined, and the instantaneous change rate of the time series energy corresponding to the data points at the preset proportion position is used as the preset filtering threshold.

5. A denoising method for partial discharge detection according to claim 1, characterized in that: In the instantaneous change rate curve of the time series energy after the first denoising, a data point where the instantaneous change rate of the time series energy is greater than zero is determined, and based on the data point where the instantaneous change rate of the time series energy is greater than zero, a pulse waveform is extracted from the high-frequency current signal to be processed, and a pulse array is constructed, specifically including: Determine the data points where the instantaneous rate of change of time series energy is greater than zero; Based on a preset judgment function, determining a first time period corresponding to a data point where the instantaneous rate of change of time series energy is zero, and setting the high-frequency current signal to be processed corresponding to the first time period to zero; Based on the preset judgment function, the second time period corresponding to the data point whose instantaneous change rate of the time series energy is greater than zero is determined, and the high-frequency current signal to be processed corresponding to the second time period is retained to obtain the pulse array.

6. A denoising method for partial discharge detection according to claim 5, characterized in that: The preset judgment function is: Among them, P(i) represents the amplitude corresponding to the i-th data point of the pulse array, x(i) represents the amplitude corresponding to the i-th data point of the high-frequency current signal to be processed; ΔE(i) represents the instantaneous change rate of the time series energy corresponding to the i-th data point.

7. A denoising method for partial discharge detection according to claim 1, characterized in that: The determining of the spectrum bandwidth corresponding to each pulse waveform in the pulse array at a preset amplitude ratio threshold, and performing secondary denoising on the high-frequency current signal to be processed based on the spectrum bandwidth to obtain a partial discharge signal specifically includes: Determine the frequency spectrum corresponding to each pulse waveform in the pulse array, and determine the maximum frequency spectrum amplitude; Determine a reference spectrum amplitude corresponding to a preset amplitude ratio threshold of the maximum spectrum amplitude; and Determining a first frequency and a second frequency corresponding to the reference spectrum amplitude; Based on the spectrum bandwidth between the first frequency and the second frequency, secondary denoising is performed on the high-frequency current signal to be processed to obtain a partial discharge signal.

8. A denoising method for partial discharge detection according to claim 7, characterized in that: The performing secondary denoising on the high-frequency current signal to be processed based on the spectrum bandwidth between the first frequency and the second frequency to obtain a partial discharge signal specifically includes: When the spectrum bandwidth between the first frequency and the second frequency is greater than a preset spectrum bandwidth threshold, determining that the pulse waveform is a partial discharge signal and performing retention processing; When the spectrum bandwidth between the first frequency and the second frequency is not greater than the preset spectrum bandwidth threshold, the pulse waveform is determined to be an interference pulse, and the amplitude of the interference pulse is set to zero.

9. A denoising device for partial discharge detection, characterized in that: The device comprises a memory for storing computer program instructions and a processor for executing the program instructions, wherein when the computer program instructions are executed by the processor, the device is triggered to execute the method according to any one of claims 1 to 8.

10. A non-volatile computer storage medium storing computer executable instructions, characterized in that: The computer executable instructions can execute the method according to any one of claims 1 to 8.

Images