A partial discharge interference suppression method based on pulse period consistency evaluation

Abstract

This invention discloses a partial discharge interference suppression method based on pulse period consistency evaluation. First, the partial discharge signal is analyzed using a cumulative energy function to calculate the average energy. Em Then, a cumulative energy curve is constructed, and after smoothing and filtering, local minima are identified as pulse start points and local maxima as pulse end points, allowing for accurate extraction of pulse timestamps. Next, periodic consistency is evaluated based on the pulse time series: the interference period is estimated using a third-order autocorrelation function. T Then, time deviation score and energy consistency score are calculated separately to quantify the periodicity characteristics of the pulse through dual scoring. Finally, pulses with scores all above the threshold are identified as periodic interference, and the partial discharge quantity of their corresponding time period is set to zero. This invention does not require changes to the hardware connection, can adaptively distinguish and eliminate periodic interference generated by power electronic devices, improve the accuracy and reliability of partial discharge monitoring, and overcome the shortcomings of traditional manual window selection methods.

Description

Technical Field

[0001] This invention belongs to the field of power equipment condition monitoring technology, specifically relating to a method for suppressing partial discharge interference based on pulse period consistency evaluation. Background Technology

[0002] In power systems, critical equipment such as switchgear must maintain safe and stable operation over the long term. However, defects such as metal burrs, suspended particles, insulation surface contamination, and poor assembly are difficult to completely eliminate during manufacturing, transportation, installation, and commissioning. These defects can easily cause distortion of the electric field distribution and trigger partial discharge. Partial discharge continuously weakens insulation strength, and if these defects evolve and intensify, they may lead to major equipment failures, resulting in widespread power outages and significant economic and safety risks.

[0003] To address the aforementioned issues, online partial discharge detection has become a common method for assessing the condition of live insulation. Common technical approaches include capacitive partial discharge detection, ultra-high frequency (UHF) methods, ultrasonic methods, and high-frequency current methods. Among these, capacitive partial discharge detection offers advantages such as high sensitivity and the ability to quantitatively measure apparent discharge. Typically, this method has a detection sensitivity of approximately 5 pC, and under ideal environmental conditions, the noise level can be reduced to 1 pC.

[0004] Existing capacitive partial discharge detection methods are mainly designed for detecting discharge pulses in the frequency band below one megahertz. However, in the low-frequency range, the detection circuit is susceptible to multi-path interference from spatial electromagnetic coupling, corona discharge, high-voltage leads, and grounding leads, which limits the effective identification of small-amplitude discharge signals.

[0005] To improve the engineering applicability of capacitive partial discharge detection methods, it is necessary to systematically study and suppress noise interference in partial discharge signals. On one hand, this can be achieved by addressing the source and propagation path, identifying interference sources and their coupling channels, and implementing measures such as isolation, shielding, and grounding optimization to reduce noise. However, the connection relationships of power equipment are usually not easily altered, limiting the practical application of such methods in the field. On the other hand, digital signal processing technology can be used to denoise and enhance the features of the acquired signals without changing the primary system connections. This approach requires less hardware modification, is highly adaptable, and is therefore easier to promote and apply in practical operation and maintenance scenarios.

[0006] Interference is typically categorized into periodic interference, random impulse interference, and white noise based on its time-domain characteristics. White noise primarily originates from thermal noise and shot noise from various devices, exhibiting a continuous and stable distribution across the frequency spectrum. Periodic interference can be further divided into periodic narrowband interference and periodic impulse interference. Periodic impulse interference is mainly caused by sudden changes in voltage and current due to the interaction between capacitance and inductance during the switching process of switching devices such as thyristors and IGBTs, generating high-energy, wideband impulse interference. This impulse signal manifests as a short-period pulse in the time domain and as a wideband signal containing multiple frequency components in the frequency domain. The waveform of impulse interference is similar to that of partial discharge pulse signals, which may lead to missed detections or false alarms in partial discharge monitoring, significantly affecting the accuracy of online partial discharge detection.

[0007] To suppress periodic pulse interference, considering that its phase is relatively fixed within the power frequency cycle, manual window selection methods are often used in the field, including direct time-axis domain and phase domain processing. However, these methods are highly subjective, require extensive experience, and may result in signal loss during the window selection process, affecting the accuracy of partial discharge detection. Summary of the Invention

[0008] To address the shortcomings of existing technologies, such as difficulty in suppressing low-frequency interference and susceptibility to subjective distortion in window selection, this invention provides a partial discharge interference suppression method for capacitive switchgear, which can suppress periodic pulse interference signals in capacitive partial discharge detection and improve detection accuracy.

[0009] To achieve the above objectives, the technical solution of the present invention is as follows: A partial discharge interference suppression method based on pulse period consistency evaluation includes the following steps: S1. Pulse timestamp extraction: Extract the start and end timestamps of each pulse from the partial discharge signal; S2. Pulse period consistency evaluation: Based on the timestamp obtained in step S1, first estimate the periodic pulse interference period, and then calculate the pulse time deviation consistency score and pulse energy consistency score. S3. Elimination of periodic pulse interference: Identify periodic pulse interference based on the scoring results of step S2, and set the partial discharge quantity of the corresponding time period to zero.

[0010] Furthermore, in step S1, the specific process of extracting the pulse timestamp includes the following steps: S11. Calculate the average energy of the entire signal segment. E m Specifically: ; Where k is the sampling point index of “partial discharge signal x[1], x[2], ..., x[N]” (k=1,2···,N), N is the maximum value of the sampling point index of a single packet of data, and x(k) is the signal amplitude corresponding to the kth sampling point; S12. Construct the cumulative energy function E(k), specifically: ; S13. Perform smoothing filtering on the cumulative energy function, identify the local minimum value of the smoothed cumulative energy function as the pulse start timestamp, and the local maximum value as the pulse end timestamp.

[0011] Furthermore, a typical characteristic of periodic pulse interference is that it occurs within one power frequency cycle. T Within 0, its phase is relatively fixed. Taking the widely used 12-pulse rectifier circuit as an example, the 12-pulse interference it generates occurs within one power frequency cycle. T It appears approximately 12 times at equal intervals within 0.

[0012] In contrast, partial discharge signals are highly random and non-periodic, exhibiting different probability distributions at different phases. Moreover, compared to partial discharge pulses, periodic pulse interference not only shows periodicity in its occurrence time but also generates relatively small fluctuations in constant pulse energy within each time period.

[0013] Based on this, a periodicity score is proposed to measure the periodicity of pulses, and its calculation method is as follows: Periodic pulse interference cycle T Sure: Taking the widely used 12-pulse rectifier circuit as an example, the 12-pulse interference it generates occurs within one power frequency cycle. T The periodic pulse interference occurs approximately 12 times at equal intervals within 0, therefore the periodic pulse interference period is... T Power frequency cycle T 0 / 12. However, in actual monitoring, the periodic pulse interference period in the real signal... T Since the period is unknown, it is necessary to estimate the period of the periodic pulse by analyzing the signal.

[0014] Since periodic impulse interference is characterized by repeating within a certain time period, these time interval sequences can be used to estimate the period of periodic impulse interference. T In order to accurately estimate the cycle T The method will use a higher-order autocorrelation function, which can effectively capture the periodic components in the signal. It is especially suitable for signals that contain both periodic pulse interference and highly random non-periodic signals.

[0015] Therefore, the estimated periodic pulse interference period specifically includes the following: S211. The start time of each pulse in the time series has been determined and is denoted as... t 11 , t 21 , t 31 ... t i1 Let the start time of the signal be... t 0, calculate each pulse relative to t A time difference of 0, i.e., constructing a time interval sequence of Δ t i ,in ; S212. Stamp the pulse start time interval sequence Δ t i Converted into point process signal S(t) Specifically: ; in, t For time; S213. Calculate the third-order autocorrelation function of the process signal at this point. R 3( τ ): , in i , j , k Indices for the start times of the i-th, j-th, and k-th pulses; Δ t i Δ is the time difference between the start time of the i-th pulse and the signal start time t0; t j Δ is the time difference between the start time of the j-th pulse and the signal start time t0; t k It is the time difference between the start time of the kth pulse and the signal start time t0; τ An intermediate variable introduced during the process of solving for period T; By searching R 3( τ The peak value determines the period of the periodic pulse interference. T， If there are pulses in the time interval sequence Δti, check if there are pulses near Δti+τ and Δti+2τ. If so, add 1 to the contribution value of τ. If τ is equal to the period T of the periodic pulse interference signal, such triple overlap will occur frequently, and the function R3(τ) will have a significant peak at τ=T, thus determining the period T of the periodic pulse signal.

[0016] Furthermore, the pulse time deviation consistency score Determined according to the following formula: , in, γ is the attenuation factor, with a value of 1; round(·) is the rounding function.

[0017] Furthermore, the calculation process for the pulse energy consistency score includes: S231. Collect partial discharge signals for 50 power frequency cycles, and construct a phase-resolved pulse sequence spectrum from the partial discharge signals containing pulses; S232. Calculate the first i The variance of 50 signals within a single phase window: , in, E j It is the first i Inside the window j The energy of a pulse; E avg It is the first i The average energy of all pulses within the phase window is calculated using the following formula: ; S233. Calculate the pulse energy consistency score, specifically as follows: , in, β This is the attenuation factor, with a value of 1.

[0018] Furthermore, the specific criterion for eliminating periodic pulse interference in step S3 is as follows: when the... i Pulse energy consistency score for each phase window S time,i The value is higher than the set threshold, and the first value within the phase window is... j Consistency score of time deviation of each pulse S energy,i If the value exceeds the set threshold, the pulse is determined to be a periodic pulse interference signal, and the partial discharge quantity during the time period from the pulse start point to the pulse end point is set to "0" pC.

[0019] Compared with the prior art, the present invention has the following advantages and beneficial effects: The partial discharge interference suppression method based on pulse period consistency evaluation proposed in this invention can effectively suppress periodic pulse interference in pulse current signals and improve the accuracy of partial discharge monitoring. Through techniques such as period consistency scoring and time deviation consistency scoring, it can accurately distinguish between periodic pulse interference and partial discharge pulse signals, and suppress interference without changing the primary system connection. This method has strong adaptability and can automatically identify and eliminate periodic interference, avoiding the subjectivity and signal loss problems of traditional manual window selection methods. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the time-domain waveform of a partial discharge signal containing pulses.

[0021] Figure 2 For cumulative energy function E(k) curve.

[0022] Figure 3 This is a partial discharge PRPS spectrum.

[0023] Figure 4 It serves as a partial discharge experimental platform.

[0024] Figure 5 PRPS spectrum when undischarged and using unshielded wire.

[0025] Figure 6 PRPS spectrum when no discharge is detected and a shielded wire is used.

[0026] Figure 7 Comparison of the effects before and after the algorithm is applied. Detailed Implementation

[0027] The invention will be further described below with reference to the accompanying drawings.

[0028] Example 1: A partial discharge interference suppression method based on pulse period consistency evaluation, comprising the following steps: S1. Pulse Timestamp Extraction: In order to analyze individual pulses in the time series of partial discharge signals, it is first necessary to extract the timestamp of each pulse, specifically the start and end timestamps of each pulse from the partial discharge signal. S11. The partial discharge signal containing the pulse... x [1], x [2],…, x [ N ], where n is the index of the sampling point ( k=1, 2···，N ( ), draw a schematic diagram of the time-domain waveform of the partial discharge signal containing pulses, such as Figure 1As shown; compared with the "no-pulse" interval, the non-stationary pulse signal exhibits a high concentration of energy in the time domain in the "pulse-containing" interval; that is, in the "no-pulse" interval, the instantaneous energy... x 2 [ k The expected value is approximately equal to the average energy of the entire signal. E m ; Calculate the average energy Em of the entire signal segment, specifically as follows: 1-1; Where k is the sampling point index of “partial discharge signal x[1], x[2], ..., x[N]” (k=1,2···,N), N is the maximum value of the sampling point index of a single packet of data, and x(k) is the signal amplitude corresponding to the kth sampling point; S12. Construct the cumulative energy function curve, such as Figure 2 As shown, the cumulative energy function E [ k It will fluctuate slightly around a near-horizontal line; once it enters the "pulse-containing" range, the instantaneous energy... x 2 [ k >> E m Cumulative energy function E ( k The pulse will then clearly "turn upwards," forming an ascending segment; therefore, the local minimum value before the start of the ascending segment is the pulse start time; construct the cumulative energy function. E ( k ), cumulative energy function E ( k Specifically: 1-2; S13. Real signals contain white noise, periodic interference, etc., which cannot be directly calculated. E [ k There will be slight fluctuations, and many false extreme points will appear. Therefore, it is also necessary to adjust the cumulative energy function. E ( k Perform another smoothing filter to suppress noise, resulting in the smoothed cumulative energy function. E ( k This makes the "valley-slope" structure truly caused by pulses clearer. When calculating the cumulative energy function of a signal containing discharge pulses and pulse interference, the cumulative energy function is not smooth and exhibits multiple extreme points, making it difficult to determine which correspond to the pulse signals. Therefore, it is necessary to apply a smoothing filter to the obtained cumulative energy function curve to smooth the cumulative energy function, identify the local minimum value of the smoothed cumulative energy function as the pulse start timestamp, and the local maximum value as the pulse end timestamp. S2. Pulse Period Consistency Evaluation: Based on the timestamp obtained in step S1, the following steps are performed: S21. First, estimate the periodic pulse interference period, specifically including the following: S211. The start time of each pulse in the time series has been determined and is denoted as... t 11 , t 21 , t 31 ... t n1 Let the start time of the signal be... t 0, calculate each pulse relative to t A time difference of 0, i.e., constructing a time interval sequence of Δ t i ,in ; S212. Partial discharge signal x(t) is processed point by point, that is, the waveform and amplitude of each pulse are no longer considered, but only as an "event" that occurs at certain moments; thus, the entire signal becomes "a series of time points of events", that is, the partial discharge signal. x ( t )exist t =Δ t i There is one pulse at any given time, therefore the pulse start time interval sequence is stamped Δ. t i Converted into point process signal S(t) Specifically: 2-1; in, t For time; S213. Calculate the third-order autocorrelation function of the process signal at this point. R 3( τ ): 2-2, in, i , j , k Indices for the start times of the i-th, j-th, and k-th pulses; Δ t iΔ is the time difference between the start time of the i-th pulse and the signal start time t0; t j Δ is the time difference between the start time of the j-th pulse and the signal start time t0; t k It is the time difference between the start time of the kth pulse and the signal start time t0; τ An intermediate variable introduced during the process of solving for period T; In the time interval sequence Δ t i Check its value in Δ t i + τ and Δ t i +2 τ Are there any pulses nearby? If so, then... τ Contribution value incremented by 1; if τ Equal to the period of the periodic pulse interference signal T Then such triple overlaps will occur frequently, and the function R 3( τ )exist τ = T A significant peak appears at a certain point, which indicates the periodicity of the pulse signal. T The certainty, therefore by searching R 3 (τ) The peak value determines the period of the periodic pulse interference. T ; S22. Calculate the pulse time deviation consistency score using the formula; If a pulse signal is a periodic pulse interference signal, then |Δ t i - round(Δ t i / T ) / T If the score is close to 0, the time bias score will be... S time,i Approaching 1; if a pulse signal is a partial discharge pulse signal, then |Δ t i - round(Δ t i / T ) / T If it's large, then the time deviation score will be... S time,i It approaches 0; therefore, since periodic pulse interference exhibits periodicity, Δ t i It will approach the periodic pulse interference period.T Integer multiples of; Δ t i With periodic pulse interference period T The deviation Δ between i For |Δ t i - round(Δ t i / T ) / T Based on this, the time deviation score is calculated using the following formula: 2-3, in γ The decay factor is 1; round(·) is the rounding function. S23. Pulse Energy Consistency Scoring: Compared to partial discharge pulse discharge, periodic pulse interference not only exhibits periodicity but also produces a constant pulse energy fluctuation with smaller fluctuations within each cycle. Therefore, the energy variance of partial discharge signals is usually larger, while the energy variance of periodic pulse interference signals is smaller. Based on this characteristic, a pulse energy consistency scoring criterion is proposed. The calculation process includes: S231. Acquire partial discharge signals containing pulses over 50 power frequency cycles. x [1], x [2],…, x [ N ], establish as Figure 3 The phase-resolved pulse sequence (PRPS) spectrum shown is used to correlate the pulses of a signal with the phases of a periodic signal to show the distribution of the signal at different phase positions. S232. Calculate the first i Variance of 50 signal energy values ​​within a single phase window: 2-4, in, E j It is the first i Inside the window j The energy of a pulse; E avg It is the average energy of all pulses within the i-th phase window, calculated using formula 2-5; S233. Calculate the pulse energy consistency score, specifically as follows: 2-6, in β This is the attenuation factor, with a value of 1; If the energy variation of the pulse within the phase window is small, then the variance is small, and the exponential function... S energy,i A value close to 1 indicates a high score, suggesting that the phase window may contain periodic pulses or white noise interference; if the energy variation of the pulses within the phase window is large, the variance is large, and the exponential function... S energy,i A value close to 0 indicates a low score, suggesting that the signal within the phase window may be a partial discharge signal or other random interference. S3. Elimination of periodic pulse interference: Identify periodic pulse interference based on the scoring results of step S2, and set the partial discharge quantity of its corresponding time period to zero; The specific criteria for eliminating periodic pulse interference are as follows: when the periodic pulse interference is eliminated... i Pulse energy consistency score for each phase window S energy,i The value is higher than the set threshold, and the first value within the phase window is... j Consistency score of time deviation of each pulse S energy,i If the value exceeds the set threshold, the pulse is determined to be a periodic pulse interference signal, and the partial discharge quantity during the time period from the pulse start point to the pulse end point is set to "0" pC.

[0029] Example 2: Build such a structure here Figure 4 The partial discharge test platform shown includes a test transformer control box A and a corona-free power frequency test transformer. T Artificially simulated defect B; Coupling capacitance C k It provides a circuit for the pulse current generated by artificially simulated defects, and avoids the direct action of power frequency high voltage on the detection impedance. Z d Its function; a partial discharge sensor with online monitoring capabilities. M .

[0030] In Experiment 1, the impedance detection and partial discharge sensor were connected using an unshielded cable. Figure 5 The PRPS spectrum, plotted from data uploaded by the partial discharge sensor without applying pressure to the artificially simulated defect, shows significant periodic pulse interference at approximately 150° and 310° phase windows.

[0031] In Experiment 2, the detection impedance and the partial discharge sensor were connected using a shielded cable. Figure 6 The PRPS spectrum plotted from the data uploaded by the partial discharge sensor when no artificially simulated defect was applied did not show any obvious periodic pulse interference.

[0032] To address this issue, the partial discharge interference suppression method based on pulse period consistency evaluation proposed in this scheme is adopted to improve the accuracy of partial discharge monitoring and reduce the impact of periodic pulse interference.

[0033] a. Pulse Timestamp Extraction: First, the timestamp of each pulse is extracted from the acquired pulse current signal. By calculating the cumulative energy function of the signal, the "pulse-containing" and "pulse-free" intervals in the signal are identified. In the "pulse-free" interval, the instantaneous energy fluctuation is small; while in the "pulse-containing" interval, the instantaneous energy increases sharply, forming a distinct rising segment. By smoothing the cumulative energy function, the start and end times of each pulse are identified, thereby accurately extracting the pulse timestamp.

[0034] b. Time Deviation Consistency Score: Based on the time deviation between the pulse and the period, a time deviation consistency score is calculated for each pulse. This score measures the magnitude of the deviation of each pulse signal relative to the periodic interference period. For periodic pulse interference signals, the time deviation consistency score is close to 1; while for partial discharge signals, the score is close to 0.

[0035] c. Pulse Energy Consistency Scoring: Periodic pulse interference signals exhibit smaller energy fluctuations, while partial discharge signals show larger energy fluctuations. By calculating the energy variance of the signal and combining it with the pulse energy changes within each phase window, the system can further determine the signal type. If the pulse energy change within a certain phase window is small, then that phase window may contain periodic pulses or white noise interference; if the pulse energy change is large, then it may be a partial discharge signal or other random interference.

[0036] d. If the i-th phase window is determined to be affected by periodic pulses or white noise interference based on the pulse energy consistency score, and the i-th phase window... j If a pulse signal is judged to be a periodic pulse interference signal by the time deviation consistency score, then the partial discharge quantity during the time period from the start time of the pulse to the end time of the pulse is set to "0" pC.

[0037] In Experiment 3, the detection impedance and the partial discharge sensor were connected using an unshielded wire, and the periodic pulse interference suppression algorithm was not implemented. Figure 7 The left side shows the PRPS spectrum plotted from the data uploaded by the partial discharge sensor when the artificially simulated defect is stressed. The discharge signal is obvious, and there are obvious periodic pulse interferences at approximately 150° and 310° phase windows.

[0038] Experiment 4: The detection impedance and the partial discharge sensor are connected with a "shielded wire" to implement a periodic pulse interference suppression algorithm. Figure 7The right side shows the PRPS spectrum plotted from the data uploaded by the partial discharge sensor when pressure is applied to the artificially simulated defect. The discharge signal is obvious, and the partial discharge data can be effectively processed by the algorithm proposed in this scheme to avoid the influence of periodic pulse interference.

[0039] In summary, this embodiment systematically verifies the effectiveness of the proposed method through a rigorously designed comparative experiment. The experiment first established a standard partial discharge test platform. The unshielded wire connection in Experiment 1 clearly revealed the inherent periodic pulse interference at the 150° and 310° power frequency phases, providing a clear target for the algorithm processing.

[0040] Subsequently, the shielded wire connection in Experiment 2 demonstrated that while hardware measures could suppress interference, they could not eliminate it completely. To highlight the core value of the algorithm, a key comparison was made: the unshielded wire in Experiment 3, without the algorithm, showed that the actual discharge and interference were severely mixed; while the shielded wire in Experiment 4, by applying the algorithm, significantly eliminated the periodic interference, making the actual discharge signal clearly visible.

[0041] The empirical process clearly demonstrates the three core steps of the algorithm: accurately extracting pulse timestamps using the cumulative energy function; estimating the interference period using the third-order autocorrelation function and quantifying the pulse periodicity characteristics by combining time and energy dual consistency scores; and finally, zeroing out pulse periods identified as interference. This embodiment powerfully proves that this method can not only effectively overcome hardware limitations but also adaptively and accurately suppress interference, providing a reliable technical guarantee for the accurate assessment of switchgear insulation status.

[0042] Numerous specific details are set forth in this specification. However, it will be understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0043] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A method for suppressing partial discharge interference based on pulse period consistency evaluation, characterized in that, Includes the following steps: S1. Pulse timestamp extraction: Extract the start and end timestamps of each pulse from the partial discharge signal; S2. Pulse period consistency evaluation: Based on the timestamp obtained in step S1, first estimate the periodic pulse interference period, and then calculate the pulse time deviation consistency score and pulse energy consistency score. S3. Elimination of periodic pulse interference: Identify periodic pulse interference based on the scoring results of step S2, and set the partial discharge quantity of the corresponding time period to zero.

2. The partial discharge interference suppression method according to claim 1, characterized in that, In step S1, the specific process of extracting the pulse timestamp includes the following steps: S11. Calculate the average energy Em of the entire signal segment, specifically as follows: ； Where k is the sampling point index of "partial discharge signal x[1], x[2], ..., x[N]" (k=1,2···,N), N is the maximum value of the sampling point index of a single packet of data, and x(k) is the signal amplitude corresponding to the kth sampling point; S12. Construct the cumulative energy function E(k), specifically: ； S13. Perform smoothing filtering on the cumulative energy function, identify the local minimum value of the smoothed cumulative energy function as the pulse start timestamp, and the local maximum value as the pulse end timestamp.

3. The partial discharge interference suppression method according to claim 1, characterized in that, In step S2, the estimation of the periodic pulse interference period specifically includes the following: S211. The start time of each pulse in the time series has been determined and is denoted as... t 11 , t 21 , t 31 ... t i1 Let the start time of the signal be... t 0, calculate each pulse relative to t A time difference of 0, i.e., constructing a time interval sequence of Δ t i ,in ; S212. Stamp the pulse start time interval sequence Δ t i Converted into point process signal S(t) Specifically: ； in, t For time; S213. Calculate the third-order autocorrelation function of the process signal at this point. R 3( τ ): ， in, i , j , k Indices for the start times of the i-th, j-th, and k-th pulses; Δ t i Δ is the time difference between the start time of the i-th pulse and the signal start time t0; t j Δ is the time difference between the start time of the j-th pulse and the signal start time t0; t k It is the time difference between the start time of the kth pulse and the signal start time t0; τ An intermediate variable introduced during the process of solving for period T; By searching R 3( τ The peak value of the periodic pulse interference determines the period of the periodic pulse interference. If there is a pulse in the time interval sequence Δti, check if there are also pulses near Δti+τ and Δti+2τ. If so, increment the contribution value of τ by 1. If τ is equal to the period T of the periodic pulse interference signal, such triple overlap will occur frequently, and the function R3(τ) will have a significant peak value at τ=T, thus determining the period T of the periodic pulse signal. T .

4. The partial discharge interference suppression method according to claim 3, characterized in that, In step S2, the pulse time deviation consistency score is... Determined according to the following formula: ， in, γ is the attenuation factor, with a value of 1; round(·) is the rounding function.

5. The partial discharge interference suppression method according to claim 1, characterized in that, In step S2, the calculation process of the pulse energy consistency score includes: S231. Collect partial discharge signals for 50 power frequency cycles, and construct a phase-resolved pulse sequence spectrum from the partial discharge signals containing pulses; S232. Calculate the first i The variance of 50 signals within a single phase window: ， in, E j It is the first i Inside the window j The energy of a pulse; E avg It is the average energy of all pulses within the i-th phase window, calculated using the following formula: ; S233. Calculate the pulse energy consistency score Specifically: ， in, β This is the attenuation factor, with a value of 1.

6. The partial discharge interference suppression method according to claim 5, characterized in that, In step S3, the specific criterion for eliminating the periodic pulse interference is: when the... i Pulse energy consistency score for each phase window S time,i The value is higher than the set threshold, and the first value within the phase window is... j Consistency score of time deviation of each pulse S energy,i If the value exceeds the set threshold, the pulse is determined to be a periodic pulse interference signal, and the partial discharge quantity during the time period from the pulse start point to the pulse end point is set to "0" pC.

Images