Method and apparatus for controlling partial discharge on the basis of active noise reduction and wavelet denoising

Abstract

The present application provides a method for controlling partial discharge on the basis of active noise reduction and wavelet denoising. The method comprises: acquiring an interface acoustic wave signal generated by breakdown of an internal cavity of a grounding resistance set apparatus; acquiring an on-site noise signal of the grounding resistance set apparatus, and performing correction of an amplitude and a phase; performing discrete processing on the interface acoustic wave signal, and performing discrete processing on the on-site noise signal of which the amplitude and phase have been corrected; performing differentiation on the discretely processed signals and performing filtering processing, to obtain a filtered acoustic wave signal; on the basis of a frequency range generated by partial discharge, acquiring the filtered acoustic wave signal, and obtaining an acoustic wave signal within a set frequency range; performing wavelet denoising processing on the acoustic wave signal within the set frequency range, and performing fitting on a waveform of the denoised acoustic wave signal within a power frequency cycle, to determine partial discharge. In the present application, by means of a technique of offsetting noise by generating an acoustic wave having a phase opposite to that of the noise, intelligent operation of the grounding resistance set apparatus is implemented.

Description

Partial discharge control method and device based on active noise reduction and wavelet denoising Technical Field

[0001] This application relates to the field of online monitoring technology for complete sets of grounding resistance devices, specifically to a partial discharge control method and device based on active noise reduction and wavelet denoising. Background Technology

[0002] Grounding resistor systems are crucial components in power supply and distribution networks. By creating a neutral point through a grounding transformer, a resistor is connected in series between the neutral point and ground. In the event of a single-phase ground fault, this resistor limits the fault current, ensuring the stable operation of the power system. However, primary equipment such as grounding transformers and primary cables can experience insulation aging and damage over time, leading to insulation defects, partial discharge, and deterioration of the grounding resistor system, ultimately affecting the stable operation of the power system.

[0003] The physical phenomena during partial discharge in power equipment are complex and diverse. Based on monitoring methods, they can be divided into electrical signal detection and non-electrical signal detection. Ultrasonic detection is a type of non-electrical signal detection, which collects acoustic signals during the partial discharge process to determine whether insulation has deteriorated. It has advantages such as no electrical contact between the detection equipment and the electrical equipment, strong anti-interference ability, and convenient and easy monitoring.

[0004] However, the acoustic signal is weak during partial discharge, and there is a lot of acoustic interference and background noise at the commissioning site of the grounding resistance system. The time spectrum and frequency spectrum of the acoustic signal directly picked up by the ultrasonic sensor are complex, and the online monitoring equipment cannot directly and effectively capture characteristic data for fault diagnosis, which brings technical difficulties to the online ultrasonic monitoring of partial discharge. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the purpose of this application is to provide a partial discharge control method and device based on active noise reduction and wavelet denoising. This method uses a technique that generates sound waves with the opposite phase to the noise to cancel it out, and obtains a relatively pure partial discharge ultrasonic signal through digital filtering and noise reduction processing. This allows for fault identification of the fault sound wave signal and facilitates the intelligent operation of the grounding resistance complete set of equipment.

[0006] One aspect of this application provides a partial discharge control method based on active noise reduction and wavelet denoising, comprising:

[0007] Acoustic wave signals generated by the breakdown of the internal cavity of the grounding resistance complete set of equipment were collected.

[0008] Collect the on-site noise signal of the grounding resistance complete set of equipment and perform amplitude and phase correction;

[0009] The collected interface acoustic wave signal is discretized, and the amplitude and phase corrected field noise signal is discretized. The discretized interface acoustic wave signal and the field noise signal are subtracted and filtered to obtain the filtered acoustic wave signal.

[0010] The filtered acoustic signal is collected according to the frequency range generated by partial discharge to obtain an acoustic signal within a set frequency range;

[0011] Wavelet denoising is performed on the acoustic signal within the set frequency range, and the waveform of the denoised acoustic signal is fitted within the power frequency cycle to determine partial discharge.

[0012] Further, the step of acquiring the interface acoustic wave signal generated by the breakdown of the internal cavity of the grounding resistance assembly using the first ultrasonic sensor and calculating the phase change, and acquiring the on-site noise signal of the grounding resistance assembly using the second ultrasonic sensor and performing amplitude and phase correction, includes:

[0013] Acquire the interface acoustic wave signal collected by the first ultrasonic sensor and the on-site noise signal collected by the second ultrasonic sensor;

[0014] Based on Snell's law, calculate the interface acoustic wave signal and phase change acquired by the first ultrasonic sensor;

[0015] Based on the spatial structure of the electrical equipment, the amplitude and phase of the on-site noise signal collected by the second ultrasonic sensor are corrected.

[0016] Furthermore, the formula for calculating the interface acoustic wave signal and phase change acquired by the first ultrasonic sensor is as follows:

[0017] In the formula, c air c is the speed of sound in air. medm θ is the speed of sound in the insulating medium; θ1 and θ2 are the angle of incidence and the angle of refraction, respectively.

[0018] Further, the step of discretizing the acquired interface acoustic wave signal, and discretizing the acquired field noise signal after amplitude and phase correction; and then subtracting and filtering the discretized acoustic wave signal to obtain a filtered acoustic wave signal, includes:

[0019] The collected interface acoustic wave signal is discretized to obtain the first discrete numerical point.

[0020] The collected field noise signal, after amplitude and phase correction, is discretized to obtain a second discrete numerical point.

[0021] Based on the first discrete numerical point and the second discrete numerical point, the first discrete numerical point and the second discrete numerical point are subtracted point by point at the same time to obtain the differenced sound wave signal.

[0022] The differenced acoustic signal is then filtered by a bandpass filter to obtain the filtered acoustic signal.

[0023] Furthermore, the transfer function of the bandpass filter is a combination of low-pass and high-pass filters, and is the bandpass Butterworth filter H(s), with the following formula:

[0024] In the formula, wc1 is the lower limit angular frequency of the passband edge; wc2 is the upper limit angular frequency of the passband edge; wn1 is the natural frequency; Q is the quality factor; and S is the Laplace operator.

[0025] Further, the wavelet denoising process performed on the acoustic signal within the set frequency range includes:

[0026] Acquire acoustic wave signals within a set frequency range, and perform FFT transformation on the acoustic wave signals within the set frequency range;

[0027] The FFT-transformed signal is then subjected to wavelet transform and multi-scale decomposition to obtain the multi-scale decomposed wavelet coefficients.

[0028] The wavelet coefficients of the multi-scale decomposition are denoised to generate denoised wavelet coefficients.

[0029] The denoised wavelet coefficients are reconstructed using wavelet transform to generate a denoised acoustic signal.

[0030] Further, the step of performing wavelet transform on the signal after FFT transformation and performing multi-scale decomposition to obtain the wavelet coefficients of multi-scale decomposition includes:

[0031] Acquire acoustic wave signals within the set frequency range, perform signal conversion, and form digital signals;

[0032] Based on the data signal, discrete wavelet transform is performed according to the set number of decomposition levels to obtain the wavelet coefficients cj of multi-scale decomposition;

[0033] The set number of decomposition layers is 5;

[0034] The wavelet coefficients cj: cj=∫f(t)ψ * (t-2 j T)dt;

[0035] In the formula: f(t) is a function with t as the variable; ψ is the wavelet function; j is the number of decomposition levels; t is the time variable; T is a constant related to the time scale.

[0036] Further, the denoising process performed on the wavelet coefficients of the multi-scale decomposition to generate denoised wavelet coefficients includes:

[0037] Obtain the wavelet coefficients of the multi-scale decomposition, and process the wavelet coefficients using Minimaxi threshold selection;

[0038] Based on the wavelet coefficients processed by the threshold, an inverse discrete wavelet transform is performed to obtain the first reconstructed signal f′(t);

[0039] The first reconstructed signal f′(t) = ∑cjψ(t-2jT);

[0040] In the formula, j is the number of decomposition levels; ψ is the wavelet function; t is the time variable; T is the constant related to the time scale; and cj is the wavelet coefficient.

[0041] Based on the first reconstructed signal f′(t), wavelet packet decomposition is performed to obtain the wavelet packet coefficients d at a predetermined decomposition level. k ;

[0042] The wavelet packet coefficients d k =∫f(t)φ * (t-2 k T)dt;

[0043] In the formula, f(t) is a function with t as the variable, φ is the wavelet packet function; t is the time variable; T is a constant related to the time scale; and k is the set number of decomposition levels.

[0044] The step of reconstructing the denoised wavelet coefficients using wavelet transform to generate a denoised acoustic signal includes:

[0045] Based on the wavelet packet coefficients, wavelet packet reconstruction is performed using wavelet transform to obtain the second reconstructed signal f′(t), generating the denoised acoustic signal:

[0046] f'(t)=∑d k φ(t-2 k T)dt;

[0047] In the formula, d k φ is the wavelet packet coefficient; t is the time variable; T is a constant related to the time scale; k is the set number of decomposition levels.

[0048] Furthermore, the step of fitting the waveform of the denoised acoustic signal within the power frequency cycle to determine partial discharge includes:

[0049] The denoised acoustic signal is acquired, the waveform of the denoised acoustic signal is sampled, and the sampled waveform is fitted.

[0050] Based on the exponential decay model of partial discharge in IEC60270, the fitted waveform is analyzed to extract characteristic parameters related to partial discharge.

[0051] The extracted feature parameters are compared with the thresholds specified in the IEC 60270 standard to determine whether partial discharge exists.

[0052] A second aspect of this application provides a partial discharge control device based on active noise reduction and wavelet denoising, comprising:

[0053] The first processing module is used to collect the interface acoustic wave signal generated by the breakdown of the internal cavity of the grounding resistance complete set of equipment.

[0054] The second processing module is used to collect the field noise signal of the grounding resistance complete set of equipment and perform amplitude and phase correction.

[0055] The calculation module is used to discretize the acquired interface acoustic wave signal, discretize the amplitude and phase corrected field noise signal, subtract the discretized interface acoustic wave signal and the field noise signal, and perform filtering to obtain the filtered acoustic wave signal.

[0056] The acquisition module is used to acquire the filtered acoustic signal according to the frequency range generated by partial discharge, and obtain the acoustic signal within the set frequency range;

[0057] The judgment module is used to perform wavelet denoising processing on the acoustic signal within the set frequency range, and to fit the waveform of the denoised acoustic signal within the power frequency cycle to judge partial discharge.

[0058] Compared with the prior art, this application has at least one of the following beneficial effects:

[0059] 1. This application effectively removes noise components from the sound wave signal by performing wavelet denoising processing on the denoised signal. During the processing, wavelet transform can decompose the signal into multiple scales, and can complete the denoising processing of a large number of sound wave signals in a short time, thereby revealing the characteristics of the signal at different scales in greater detail and improving the quality of the signal.

[0060] 2. This application, by collecting and analyzing interface acoustic wave signals and on-site noise signals, can effectively distinguish between the acoustic waves generated by partial discharge and background noise, promptly detect partial discharge phenomena inside the grounding resistance complete set of equipment, effectively distinguish between interface acoustic wave signals generated by cavity breakdown inside the grounding resistance complete set of equipment and on-site noise signals, and obtain relatively pure partial discharge ultrasonic signals through digital filtering and noise reduction processing, so as to identify fault acoustic wave signals, facilitate the intelligent operation of the grounding resistance complete set of equipment, realize the status monitoring and fault diagnosis of the grounding resistance complete set of equipment, and improve the accuracy of partial discharge detection. Attached Figure Description

[0061] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0062] Figure 1 is a flowchart of a partial discharge control method based on active noise reduction and wavelet denoising in one embodiment of this application.

[0063] Figure 2 is a flowchart of wavelet denoising processing in one embodiment of this application.

[0064] Figure 3 is a structural diagram of a partial discharge control device based on active noise reduction and wavelet denoising according to an embodiment of this application.

[0065] Figure 4 is a discrete control flowchart in one embodiment of this application.

[0066] Figure 5 is a waveform diagram of the sound source collected in one embodiment of this application.

[0067] Figure 6 is a waveform diagram of the sound source after noise reduction in one embodiment of this application.

[0068] Figure 7 is a waveform diagram of the sound source after noise reduction in one embodiment of this application. Detailed Implementation

[0069] The present application will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present application, but do not limit the present application in any way. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present application. These all fall within the protection scope of the present application.

[0070] Referring to Figure 1, an embodiment of this application shows a partial discharge control method based on active noise reduction and wavelet denoising, including: S1, collecting the interface acoustic wave signal generated by the breakdown of the cavity inside the grounding resistance assembly;

[0071] S2. Collect the on-site noise signal of the grounding resistance complete set of equipment and perform amplitude and phase correction.

[0072] S3. Discretize the collected interface acoustic wave signal and the field noise signal after amplitude and phase correction; subtract the discretized interface acoustic wave signal and the field noise signal and filter them to obtain the filtered acoustic wave signal.

[0073] S4. Acquire the filtered acoustic signal according to the frequency range generated by partial discharge to obtain the acoustic signal within the set frequency range.

[0074] S5. Perform wavelet denoising on the acoustic signal within the set frequency range, and fit the waveform of the denoised acoustic signal within the power frequency cycle to determine partial discharge.

[0075] This application, by collecting and analyzing interface acoustic wave signals and on-site noise signals, can effectively distinguish between the acoustic waves generated by partial discharge and background noise, promptly detect partial discharge phenomena inside the grounding resistance assembly, and effectively differentiate between interface acoustic wave signals generated by cavity breakdown inside the grounding resistance assembly and on-site noise signals. Digital filtering and noise reduction processing yields relatively pure partial discharge ultrasonic signals, enabling fault identification of fault acoustic wave signals, facilitating intelligent operation of the grounding resistance assembly, realizing status monitoring and fault diagnosis of the grounding resistance assembly, and improving the accuracy of partial discharge detection.

[0076] Specifically, firstly, the interface acoustic wave signal generated by the breakdown of the internal cavity of the grounding resistance assembly is acquired using a first ultrasonic sensor, and the noise signal at the site is acquired using a second ultrasonic sensor. Amplitude and phase corrections are applied to both the interface acoustic wave signal and the noise signal to reduce interference and errors. Next, the interface acoustic wave signal acquired by the first ultrasonic sensor is discretized, and the pre-processed noise signal acquired by the second ultrasonic sensor is also discretized. Then, the difference between the two discretized acoustic wave signals is calculated to eliminate the influence of background noise, and the differenced signal is filtered to obtain a filtered acoustic wave signal. Next, based on the frequency range of partial discharge, the filtered acoustic wave signal is acquired to obtain an acoustic wave signal within a set frequency range. Wavelet denoising processing is then applied to the acoustic wave signal within the set frequency range to further reduce noise interference. Finally, the denoised acoustic wave waveform is fitted within the power frequency cycle, and the presence of partial discharge is determined based on the characteristics of the fitted acoustic wave waveform, thus achieving intelligent operation of the grounding resistance assembly.

[0077] The phase change calculation of the interface acoustic wave signal aims to obtain the specific phase characteristics generated when the acoustic wave signal breaks down in the cavity inside the grounding resistor assembly, in order to identify partial discharge phenomena. The amplitude and phase correction of the field noise signal is intended to eliminate or reduce potential amplitude and phase differences between the noise signal and the interface acoustic wave signal, ensuring that the two signals accurately reflect the characteristics of the actual partial discharge signal during subsequent differential processing, thereby improving the accuracy and reliability of partial discharge detection.

[0078] In some possible embodiments, the process of acquiring interface acoustic wave signals generated by the breakdown of the internal cavity of the grounding resistor assembly, calculating the phase change, acquiring the field noise signal of the grounding resistor assembly, and performing amplitude and phase correction includes: acquiring interface acoustic wave signals acquired by a first ultrasonic sensor and field noise signals acquired by a second ultrasonic sensor; calculating the interface acoustic wave signals and phase changes acquired by the first ultrasonic sensor according to Snell's law; and correcting the amplitude and phase of the field noise signal acquired by the second ultrasonic sensor according to the spatial structure of the electrical equipment.

[0079] In amplitude and phase correction, the time stamps of the two sensors are first aligned. Then, according to the set value, the signal in the set frequency range is conditioned. The amplitude and phase of the first and second sensors are filtered, de-jittered, and interpolated to obtain the corrected data.

[0080] The first ultrasonic sensor collects specific acoustic signals generated when the cavity inside the grounding resistor assembly breaks down, such as the location and intensity of the discharge. It calculates the phase change of the interface acoustic signal. For subsequent signal synchronization and noise reduction processing, the phase of the noise reduction signal is accurately adjusted using the phase information of the discharge signal to align it with the original noise signal, thus achieving more effective noise reduction. The second ultrasonic sensor collects noise signals from the installation site, including environmental noise and operating noise. Since the site noise signal may be affected by various factors (such as propagation path, reflection, attenuation, etc.), its amplitude and phase are corrected according to the spatial structure of the electrical equipment to eliminate the influence of site noise on the discharge signal detection.

[0081] As shown in Figure 5, the sound source was collected on site. In this application, the location where partial discharge occurs can be regarded as the sound source. As a mechanical wave, the sound wave takes different forms during the partial discharge process. First, the sound wave is generated by the breakdown of the cavity inside the grounding resistor assembly. There is an interface between air and insulating material, which will cause refraction and reflection. The interface wave leaves the interface during propagation and enters a single medium. It can propagate again in the form of a spherical wave. The propagation of sound waves in gas and liquid is in the form of a spherical wave that spreads outwards. The energy of the sound wave will also be reduced. Based on this principle, active noise reduction technology is adopted.

[0082] In the above principle, the reflection coefficient R is:

[0083] Refractive index Z:

[0084] In the formula, ρ air c air ρ medm c medm These represent the density of air and the speed of sound in the insulating medium, respectively.

[0085] The reflection coefficient and refractive index are calculated for noise reduction processing.

[0086] According to Snell's law, the formula for the phase change of the interface sound wave acquired by the first ultrasonic sensor is:

[0087] In the formula, c air c is the speed of sound in air. medm θ is the speed of sound in the insulating medium; θ1 and θ2 are the angle of incidence and the angle of refraction, respectively.

[0088] In some possible embodiments, the acquired interface acoustic wave signal is discretized, and the acquired field noise signal, after amplitude and phase correction, is also discretized; the discretized acoustic wave signal is then subtracted and filtered to obtain a filtered acoustic wave signal, including:

[0089] The collected interface acoustic wave signal is discretized to obtain the first discrete numerical point; the collected field noise signal, after amplitude and phase correction, is discretized to obtain the second discrete numerical point.

[0090] Based on the first discrete numerical point and the second discrete numerical point, the first discrete numerical point and the second discrete numerical point are subtracted point by point at the same time to obtain the differenced sound wave signal; the differenced sound wave signal is then filtered by a bandpass filter to obtain the filtered sound wave signal.

[0091] As shown in Figure 6, the acquired sound source waveform is processed to reduce the influence of redundant noise. Specifically, the interface sound wave signal is directly acquired by the first ultrasonic sensor, and the sound wave signal is acquired by the second ultrasonic sensor. The continuous sound wave signal acquired by the first ultrasonic sensor is discretized to convert the continuous signal into a first discrete value point. The field noise signal acquired by the second ultrasonic sensor, after amplitude and phase correction, is discretized in the same way to obtain a second discrete value point. At the same time, based on time synchronization, the first and second discrete value points are subtracted point by point. For each time point, the difference between the first and second discrete value points is calculated. Then, the differenced sound wave signal is filtered by a bandpass filter to achieve the purpose of active noise reduction.

[0092] Referring to Figure 4, P(z) represents the discrete processing of the first ultrasonic sensor, and A(z) represents the discrete processing of the second ultrasonic sensor after amplitude and phase correction.

[0093] By using x(n) as the sound source collected on-site, the signal is discretized by the first ultrasonic sensor. Simultaneously, the signal after amplitude and phase correction is discretized by the second ultrasonic sensor, and the difference between the signals is calculated to remove the components corresponding to the on-site noise from the interface acoustic wave signal. This generates a compensation signal with the opposite phase and equal amplitude to the on-site noise, which cancels out the on-site noise in real time, achieving the effect of active noise reduction. Then, after BUTTER filtering, the filtered signal d(n) is obtained.

[0094] In the above embodiments, specifically, the transfer function of the bandpass filter is a combination of low-pass and high-pass filters, which is the bandpass Butterworth filter H(s), and the formula is:

[0095] In the formula, wc1 is the lower limit angular frequency of the passband edge; wc2 is the upper limit angular frequency of the passband edge; wn1 is the natural frequency; Q is the quality factor; and s is the Laplace operator.

[0096] Among them, the frequency range of the partial discharge is 50kHz to 400kHz. After noise reduction processing (i.e., the signal after being filtered by a bandpass Butterworth filter), the acoustic wave signal d(x) (i.e. d(n)) within the frequency range is collected.

[0097] Referring to Figure 2, in some possible embodiments, wavelet denoising processing is performed on the acoustic signal within a set frequency range, including: acquiring the acoustic signal within the set frequency range; performing FFT transformation on the acoustic signal within the set frequency range; performing wavelet transform on the signal after FFT transformation and performing multi-scale decomposition to obtain multi-scale decomposed wavelet coefficients; performing denoising processing on the multi-scale decomposed wavelet coefficients to generate denoised wavelet coefficients; and reconstructing the denoised wavelet coefficients using wavelet transformation to generate the denoised acoustic signal.

[0098] This application effectively removes noise components from acoustic signals by performing wavelet denoising on the denoised signals. During the processing, wavelet transform can decompose the signal into multiple scales, enabling the denoising of a large number of acoustic signals in a short time, thereby revealing the characteristics of the signal at different scales in greater detail and improving signal quality.

[0099] Specifically, firstly, a bandpass filter is used to filter the acquired acoustic signal to remove noise and interference signals outside the set frequency range. The acoustic signal within the set frequency range is then preprocessed (FFT transformation) to facilitate subsequent wavelet transform. Using wavelet transform technology, the preprocessed acoustic signal is converted from the time domain to the wavelet domain and decomposed into wavelet coefficients at different scales (or frequencies). Next, based on the wavelet coefficients obtained from the multi-scale decomposition, and using appropriate denoising strategies such as thresholding, the coefficients are denoised to remove noise components while retaining the main features of the signal. After denoising, denoised wavelet coefficients are obtained. Finally, using inverse wavelet transform technology, the denoised wavelet coefficients are converted back from the wavelet domain to the time domain to reconstruct the denoised acoustic signal.

[0100] Specifically, the process involves performing an FFT transform on the acoustic signal within a set frequency range; then performing a wavelet transform on the FFT-transformed signal and multi-scale decomposition to obtain the wavelet coefficients of the multi-scale decomposition. This includes: acquiring the acoustic signal within a set frequency range, performing signal conversion to form a data signal; and performing a discrete wavelet transform on the data signal according to a set number of decomposition levels to obtain the wavelet coefficients cj of the multi-scale decomposition.

[0101] The acoustic signal d(x) after active noise reduction is read to form a data sequence, and an FFT transformation is performed on the signal. Then, multi-scale decomposition is performed using wavelet transform to determine the number of decomposition levels. A discrete wavelet transform is then performed on f(t), where f(t) represents a function with respect to t, resulting in a series of wavelet coefficients cj. These wavelet coefficients represent the signal components at different scales (or frequencies).

[0102] The set number of decomposition layers is 5.

[0103] Wavelet coefficients cj: cj=∫f(t)ψ * (t-2 j T)dt.

[0104] In the formula: f(t) is a function with t as the variable; ψ is the wavelet function; j is the number of decomposition levels; t is the time variable; T is a constant related to the time scale.

[0105] First, through precise digital signal conversion, acoustic signals can be efficiently converted into processable data sequences. Then, by using Fourier transform, key frequency components in the signal are identified, and discrete wavelet transform is performed on the signal through a set number of decomposition levels, decomposing the signal into wavelet coefficients of multiple scales, thereby improving the accuracy and efficiency of acoustic signal processing.

[0106] Specifically, based on the wavelet coefficients of the multi-scale decomposition, the wavelet coefficients of the multi-scale decomposition are denoised to generate denoised wavelet coefficients; based on the denoised wavelet coefficients, the wavelet coefficients are reconstructed using wavelet transform to generate a denoised acoustic signal, including: obtaining the wavelet coefficients of the multi-scale decomposition, processing the wavelet coefficients using Minimaxi threshold selection; and performing inverse discrete wavelet transform based on the thresholded wavelet coefficients to obtain the first reconstructed signal f′(t), where the first reconstructed signal f′(t) = ∑cjψ(t-2jT).

[0107] In the formula, j is the number of decomposition levels; ψ is the wavelet function; t is the time variable; T is the constant related to the time scale; and cj is the wavelet coefficient.

[0108] Based on the first reconstructed signal f′(t), wavelet packet decomposition is performed to obtain the wavelet packet coefficients d at a predetermined decomposition level. k Wavelet packet coefficients d k =∫f(t)φ * (t-2 k T)dt.

[0109] In the formula, f(t) is a function with t as the variable, φ is the wavelet packet function; t is the time variable; T is a constant related to the time scale; and k is the set number of decomposition layers.

[0110] Based on the wavelet packet coefficients, wavelet packet reconstruction is performed using wavelet transform to obtain the second reconstructed signal f′(t), generating the denoised acoustic signal: f′(t)=∑d k φ(t-2 k T)dt.

[0111] In the formula, d k φ is the wavelet packet coefficient; t is the time variable; T is a constant related to the time scale; k is the set number of decomposition levels.

[0112] By employing the Minimaxi threshold selection strategy to finely process the wavelet coefficients, noise components are intelligently identified and suppressed while retaining the effective information in the signal to the maximum extent, generating denoised wavelet coefficients. Then, wavelet transform is used to perform inverse discrete wavelet transform on the thresholded wavelet coefficients to reconstruct the denoised signal f′(t), thereby restoring the original form of the signal, as shown in Figure 7.

[0113] In some specific embodiments, fitting the waveform of the denoised acoustic signal within the power frequency cycle to determine partial discharge includes: acquiring the denoised acoustic signal; sampling the waveform of the denoised acoustic signal; fitting the sampled waveform; analyzing the fitted waveform based on the exponential decay model of partial discharge in IEC60270; and comparing the extracted feature parameters with the threshold values ​​specified in the IEC60270 standard to determine whether partial discharge exists.

[0114] A second aspect of this application provides a partial discharge control device based on active noise reduction and wavelet denoising, comprising: a first processing module for acquiring interface acoustic wave signals generated by cavity breakdown inside a grounding resistor assembly; a second processing module for acquiring field noise signals from the grounding resistor assembly and performing amplitude and phase corrections; a calculation module for discretizing the acquired interface acoustic wave signals and the amplitude- and phase-corrected field noise signals; subtracting the discretized interface acoustic wave signals and the field noise signals, and then filtering them to obtain filtered acoustic wave signals; an acquisition module for acquiring the filtered acoustic wave signals according to the frequency range generated by partial discharge to obtain acoustic wave signals within a set frequency range; and a judgment module for performing wavelet denoising on the acoustic wave signals within the set frequency range and fitting the waveform of the denoised acoustic wave signals within a power frequency period to determine partial discharge.

[0115] As shown in Figure 3, the active noise reduction system compensates for environmental noise by performing open-loop signal filtering. The hardware system mainly includes a first ultrasonic sensor for signal acquisition and a second ultrasonic sensor for acquiring disturbance (environmental noise). Partial discharge generates a sound source. The first ultrasonic sensor is placed on the outer wall of the grounding resistor assembly structure and is tightly integrated with the outer wall. Through conduction, it can more accurately acquire the partial discharge sound signal. The second ultrasonic sensor is placed outside the grounding resistor assembly to acquire environmental noise. The ultrasonic signal is converted into an electrical signal by the first and second acoustic-to-electric conversion modules. Subsequently, the signal is filtered and amplified by the signal conditioning circuit. The on-site noise is complex and diverse. The signal is then sent to the processor for signal processing after passing through the AD conversion module.

[0116] The specific embodiments of this application have been described above. It should be understood that this application is not limited to the specific embodiments described above, and those skilled in the art can make various modifications or variations within the scope of the claims, which do not affect the substantive content of this application. The above-described preferred features can be used in any combination without conflict.

Claims

1. A partial discharge control method based on active noise reduction and wavelet denoising, characterized in that, The method comprises the following steps: Collecting interface acoustic wave signals generated by the breakdown of the internal cavity of the grounding resistance complete device; Collecting field noise signals of the grounding resistance complete device and performing amplitude and phase correction; Discretely processing the collected interface acoustic wave signals, discretely processing the field noise signals after amplitude and phase correction, subtracting the discretely processed interface acoustic wave signals and field noise signals, and performing filter processing to obtain filtered acoustic wave signals; Collecting the filtered acoustic wave signals according to the frequency range generated by partial discharge to obtain acoustic wave signals in the set frequency range; Performing wavelet denoising processing on the acoustic wave signals in the set frequency range, and fitting the waveform of the denoised acoustic wave signals in the power frequency cycle to judge the partial discharge.

2. The partial discharge control method based on active noise reduction and wavelet denoising according to claim 1, characterized in that, The method for collecting interface acoustic wave signals generated by the breakdown of the internal cavity of the grounding resistance complete device, collecting field noise signals of the grounding resistance complete device, and performing amplitude and phase correction comprises the following steps: Obtaining interface acoustic wave signals collected by a first ultrasonic sensor and field noise signals collected by a second ultrasonic sensor; According to Snell's law, calculating the interface acoustic wave signals collected by the first ultrasonic sensor and the phase change; According to the spatial structure of the electrical equipment, performing amplitude and phase correction on the field noise signals collected by the second ultrasonic sensor.

3. The partial discharge control method based on active noise reduction and wavelet denoising according to claim 2, characterized in that, The formula for calculating the interface acoustic wave signal collected by the first ultrasonic sensor and the phase change is: where c air is the speed of sound in air; c medm is the speed of sound in the insulating medium; θ1 and θ2 are the angles of incidence and refraction, respectively.

4. The partial discharge control method based on active noise reduction and wavelet denoising according to claim 1, characterized in that, The method for discretely processing the collected interface acoustic wave signals, discretely processing the field noise signals after amplitude and phase correction, subtracting the discretely processed interface acoustic wave signals and field noise signals, and performing filter processing to obtain filtered acoustic wave signals comprises the following steps: Discretely processing the collected interface acoustic wave signals to obtain first discrete numerical points; Discretely processing the field noise signals after amplitude and phase correction to obtain second discrete numerical points; According to the first discrete numerical points and the second discrete numerical points, subtracting the first discrete numerical points and the second discrete numerical points point by point at the same time to obtain subtracted acoustic wave signals; According to the subtracted acoustic wave signals, performing filter processing through a band-pass filter to obtain filtered acoustic wave signals.

5. The partial discharge control method based on active noise reduction and wavelet denoising according to claim 4, characterized in that, The transfer function of the bandpass filter is a combination of low pass and high pass, which is a bandpass Butterworth filter H(s), the formula is: In the formula, wc1 is the lower limit angular frequency of the passband edge, wc2 is the upper limit angular frequency of the passband edge, wn1 is the natural frequency, Q is the quality factor, and s is the Laplace operator.

6. The partial discharge control method based on active noise reduction and wavelet denoising according to claim 1, characterized in that, The wavelet denoising processing on the acoustic wave signals in the set frequency range comprises the following steps: Obtaining acoustic wave signals in the set frequency range, performing FFT change on the acoustic wave signals in the set frequency range; According to the FFT changed signals, performing wavelet transform and multi-scale decomposition to obtain multi-scale decomposed wavelet coefficients; Performing denoising processing on the multi-scale decomposed wavelet coefficients to generate denoised wavelet coefficients; Using wavelet change to reconstruct the denoised wavelet coefficients to generate denoised acoustic wave signals.

7. The partial discharge control method based on active noise reduction and wavelet denoising according to claim 6, characterized in that, The method for performing wavelet transform on the FFT changed signals and performing multi-scale decomposition to obtain multi-scale decomposed wavelet coefficients comprises the following steps: Obtaining acoustic wave signals in the set frequency range, performing signal conversion to form digital signals; According to the data signal, a discrete wavelet transform is performed according to a set decomposition layer number to obtain a multi-scale decomposed wavelet coefficient cj; The set decomposition layer number is 5 layers; The wavelet coefficient cj is: cj = f(t) ψ * (t-2 j T)dt; In the formula, f(t) is a function with t as a variable; ψ is a wavelet function; j is a decomposition layer number; t is a time variable; and T is a constant related to a time scale.

8. The partial discharge control method based on active noise reduction and wavelet denoising according to claim 6, characterized in that, The multi-scale decomposed wavelet coefficient is subjected to a denoising processing to generate a denoised wavelet coefficient, including: The multi-scale decomposed wavelet coefficient is subjected to a denoising processing to generate a denoised wavelet coefficient, including: According to the wavelet coefficient subjected to the threshold processing, an inverse discrete wavelet transform is performed to obtain a first reconstructed signal f'(t); The first reconstructed signal f'(t) = ∑cjψ(t-2jT); In the formula, j is a decomposition layer number; ψ is a wavelet function; t is a time variable; and T is a constant related to a time scale; and cj is a wavelet coefficient; According to the first reconstructed signal f'(t), wavelet packet decomposition is performed to obtain wavelet packet coefficients d of a set decomposition level k ; The wavelet packet coefficients d k =∫f(t)φ * (t-2 k T)dt; In the formula, f(t) is a function with t as a variable; φ is a wavelet packet function; t is a time variable; T is a constant related to a time scale; and k is a set decomposition layer number. The denoised wavelet coefficient is reconstructed by using a wavelet change to generate a denoised acoustic wave signal, including: According to the wavelet packet coefficient, a wavelet packet reconstruction is performed by using a wavelet change to obtain a second reconstructed signal f'(t) and generate a denoised acoustic wave signal: f'(t) = ∑d k φ(t-2 k T)dt; In the formula, d k is a wavelet packet coefficient; φ is a wavelet packet function; t is a time variable; T is a constant related to a time scale; and k is a set decomposition level.

9. The partial discharge control method based on active noise reduction and wavelet denoising according to claim 6, characterized in that, The waveform of the denoised acoustic wave signal in a power frequency cycle is fitted to determine a partial discharge, including: The waveform of the denoised acoustic wave signal is sampled, and the sampled waveform is fitted; Based on an exponential decay model of the partial discharge in IEC60270, the fitted waveform is analyzed to extract a characteristic parameter related to the partial discharge; The extracted characteristic parameter is compared with a threshold value specified in IEC60270 to determine whether the partial discharge exists.

10. A partial discharge control device based on active noise reduction, wavelet denoising, characterized in that, Including: The first processing module is used for collecting the interface acoustic wave signal generated by the cavity breakdown of the grounding resistance complete device; The second processing module is used for collecting the field noise signal of the grounding resistance complete device and performing amplitude and phase correction; The calculation module is used for performing discrete processing on the collected interface acoustic wave signal and the field noise signal subjected to the amplitude and phase correction; performing subtraction on the discrete processed interface acoustic wave signal and field noise signal; and performing filter processing to obtain a filtered acoustic wave signal; The collection module is used for collecting the filtered acoustic wave signal according to a frequency range generated by the partial discharge to obtain an acoustic wave signal in a set frequency range; The judgment module is used for performing wavelet denoising processing on the acoustic wave signal in the set frequency range and fitting the waveform of the denoised acoustic wave signal in a power frequency cycle to determine the partial discharge.

Images