Method and System for Extracting Moment Feature Parameters of Time-Frequency Image of Partial Discharge Pulse Waveform

Abstract

The present invention relates to a method and system for extracting moment characteristic parameters of time-frequency images of partial discharge pulse waveforms. The method includes: Step 101, performing wavelet denoising on a single pulse waveform in a pulse group and normalizing the amplitude to form a time-domain waveform; Step 102, performing time-frequency transformation on the time-domain waveform to form a time-frequency image; Step 103, calculating the first-order and second-order moments of time and frequency of the time-frequency energy distribution of the time-frequency image; and calculating the sum of the normalized amplitudes under the absolute value of the first-order moment of time and the second-order moment of time, and the sum of the normalized amplitudes under the absolute value of the first-order moment of frequency and the second-order moment of frequency, respectively, to obtain an equivalent time moment and an equivalent frequency moment; Step 104, extracting characteristic parameters for all pulse waveforms in the pulse group, that is, calculating the corresponding equivalent time moment and equivalent frequency moment, to obtain a plane distribution of equivalent time-frequency moment characteristic parameters. Compared with the prior art, the present invention has advantages such as stronger robustness.

Description

Technical Field

[0001] The present invention relates to partial discharge detection technology, and in particular to a method and system for extracting moment characteristic parameters of partial discharge pulse waveform time-frequency images. Background Art

[0002] Professor G.C. Montanari of the University of Bologna in Italy and others started in 2002, and Professor Li Yanming of Xi'an Jiaotong University in China and others started in 2008. They successively proposed a partial discharge (PD) pulse source detection technology based on ultra-wideband detection (the frequency band can cover or involve the conventional electrical detection (30 kHz - 1 MHz), high frequency (3 MHz - 30 MHz), and very high frequency (30 MHz - 300 MHz) defined by GB / T 7354, GB / T 20833.1, GB / T 20833.2, and GB / T 23642). Pulse waveform-time series detection is used instead of traditional pulse peak-time series detection, that is, recording a single pulse waveform and its acquisition time (which can be phase information under alternating voltage); since ultra-wideband detection retains relatively "complete" waveform information of the pulse, and the waveforms generated by different pulse sources have "self-similarity", using a certain "method" to quickly classify the acquired mixed original pulse group can achieve the separation between pulse sources, that is, achieve the separation between PD sources and noise sources or multiple PD sources.

[0003] The above-mentioned certain "method" for quickly classifying the acquired mixed original pulse group is the key to the implementation of this ultra-wideband technology, which is divided into two parts: ① is the method for extracting pulse waveform characteristic parameters; ② is the clustering technology based on the distribution of characteristic parameters. Among them, for the first part, "Fast Classification of Multiple Partial Discharge Pulse Groups Based on Waveform Nonlinear Mapping" published in the Journal of Electrical Engineering in March 2009 proposed a nonlinear mapping method for pulse waveforms. That is, the extraction results are displayed in a two-dimensional plane or three-dimensional space, and then means such as intelligent clustering analysis (the second part of the technology) are used to separate the pulse group to form sub-pulse groups with their own characteristics, so as to achieve the separation of multiple PD sources and noise sources. The enterprise standard of State Grid "Q / GDW 11400 - 2015 On-site Application Guide for High-Frequency Partial Discharge Live Testing Technology of Power Equipment" gives the Figure 1 extraction method of pulse waveform characteristic parameters such as equivalent frequency (MHz) and equivalent duration (ns) as shown, and uses corresponding instruments to carry out the application of live detection of power equipment.

[0004] However, in actual engineering applications, due to the influence of the propagation path, there are working conditions where the pulse waveforms obtained by ultra-wideband detection of noise sources or PD sources are very similar in time-domain waveform characterization. This makes Figure 1The method for extracting the characteristic parameters of the pulse waveform of equivalent frequency and equivalent duration shown cannot fully meet the requirements for separating multiple PD sources and noise sources by means of intelligent clustering analysis and other methods in the later stage. Summary of the Invention

[0005] The purpose of the present invention is to overcome the defects existing in the above-mentioned prior art and provide a method and system for extracting the moment characteristic parameters of the time-frequency image of the partial discharge pulse waveform.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] According to one aspect of the present invention, a method for extracting the moment characteristic parameters of the time-frequency image of the partial discharge pulse waveform is provided. This method is aimed at the mixed pulse waveform-time series, i.e., the pulse group, formed by the pulse source obtained through ultra-wideband detection. The method specifically includes the following steps:

[0008] Step 101: Use wavelet denoising and amplitude normalization for a single pulse waveform in the pulse group to form a time-domain waveform.

[0009] Step 102: Perform time-frequency transformation on the time-domain waveform to form a time-frequency image.

[0010] Step 103: Calculate the first-order and second-order moments of time and frequency of the time-frequency energy distribution of the time-frequency image; and calculate the sum of the normalized amplitudes under the absolute value of the first-order moment of time and the second-order moment of time, as well as the sum of the normalized amplitudes under the absolute value of the first-order moment of frequency and the second-order moment of frequency, respectively, to obtain the equivalent time moment and the equivalent frequency moment.

[0011] Step 104: Extract the characteristic parameters, i.e., obtain the corresponding equivalent time moment and equivalent frequency moment, for all pulse waveforms in the pulse group to obtain the planar distribution of the equivalent time-frequency moment characteristic parameters.

[0012] According to another aspect of the present invention, a system for the method for extracting the moment characteristic parameters of the time-frequency image of the partial discharge pulse waveform is provided, including a pulse waveform-time series module that mixes PD sources and noise sources obtained through ultra-wideband detection, a pulse time-domain waveform module, a time-frequency transformation module, a first-order time moment module, a second-order time moment module, a first-order frequency moment module, a second-order frequency moment module, an equivalent time moment module, an equivalent frequency moment module, and a planar distribution module of the equivalent time-frequency moment characteristic parameters;

[0013] The pulse waveform-time series module that mixes the PD source and the noise source obtained by the ultra-wideband detection is connected to the time-frequency transformation module through the pulse time-domain waveform module. The time-frequency transformation module is respectively connected to the first-order time moment module, the second-order time moment module, the first-order frequency moment module, and the second-order frequency moment module. The first-order time moment module and the second-order time moment module are respectively connected to the equivalent time moment module. The first-order frequency moment module and the second-order frequency moment module are respectively connected to the equivalent frequency moment module. The equivalent time moment module and the equivalent frequency moment module are respectively connected to the equivalent time-frequency moment characteristic parameter plane distribution module.

[0014] As a preferred technical solution, the pulse waveform-time series module that mixes the PD source and the noise source obtained by the ultra-wideband detection uses a data acquisition device with an analog bandwidth of dozens of MHz and a sampling rate of 100 MS / s or more, and through a coupling device with a frequency response of dozens of MHz or more that meets the nanosecond-level PD ultra-wideband detection, based on the pulse waveform triggering technology, records the single-pulse time-domain waveform and the pulse waveform-time series at the corresponding trigger moment, that is, the pulse group p j (t), which is defined as follows:

[0015]

[0016] In the formula:

[0017] j is the jth pulse, where j = 1, 2,... N, and N is the total number of pulse waveforms included in the pulse group;

[0018] k is that the pulse waveform consists of k points, and the number of points is determined by the sampling rate f s *sampling duration.

[0019] As a preferred technical solution, the system further includes a wavelet denoising module provided between the pulse waveform-time series module that mixes the PD source and the noise source obtained by the ultra-wideband detection and the time-frequency transformation module. This wavelet denoising module is a discrete dyadic wavelet, and after wavelet denoising and amplitude normalization of the single original pulse waveform in the pulse group, it forms the time-domain waveform, that is, p j (t).

[0020] As a preferred technical solution, the pulse time-domain waveform module is used to store the time-domain waveform, that is, p j (t), as the object to be processed by the subsequent modules.

[0021] As a preferred technical solution, the time-frequency transformation module uses the generalized rectangular time-frequency distribution algorithm to perform time-frequency transformation on p j (t) to obtain the tfr j (t, f) time-frequency image.

[0022] As a preferred technical solution, the first-order time moment module, the second-order time moment module, the first-order frequency moment module, and the second-order frequency moment module are defined as follows in sequence:

[0023]

[0024]

[0025]

[0026]

[0027] where f m (t) j is the first-order time moment, B 2 (t) j is the second-order time moment, t m (f) j is the first-order frequency moment, T 2 (t) j is the second-order frequency moment, and tfr j (t, f) is the time-frequency image.

[0028] As a preferred technical solution, the equivalent time moment module is calculated as follows:

[0029]

[0030] As a preferred technical solution, the equivalent frequency moment module is calculated as follows:

[0031]

[0032] As a preferred technical solution, the calculation of the equivalent time-frequency moment feature parameter plane distribution module is as follows:

[0033] TF j =(max(T j ))_N, max(F j ))_N)

[0034] where T j is the equivalent time moment and F j is the equivalent frequency moment.

[0035] Compared with the prior art, the present invention has the following advantages:

[0036] 1. The present invention makes full use of the joint time-frequency distribution characteristics of the non-stationary characterization of the pulse waveform signal in the time domain and the frequency domain, and obtains the characteristic parameters by using the time and frequency moment distributions of the time-frequency image.

[0037] 2. The present invention has stronger robustness than the current method for advancing characteristic parameters based only on the pulse time domain and frequency domain waveforms (such as the equivalent time-frequency method), and has the advantages of simple algorithm and convenient implementation.

[0038] 3. The most important thing about the present invention is that it can process pulse sources with similar time domain characteristics of pulse waveforms. The plane distribution of equivalent time-frequency moment characteristic parameters has the characteristics of "intra-class cohesion" and "inter-class separation", and is suitable for constructing a fast classification technology for pulse groups in ultra-wideband PD detection. Description of the Drawings

[0039] Figure 1 is an example of the result of the current method for extracting characteristic parameters using equivalent frequency and equivalent duration;

[0040] Figure 2 are the main modules of the method shown in the present invention;

[0041] Figure 3 is the processing flow chart of pulse data in the method of the present invention;

[0042] Figure 4 is the original waveform of a typical single PD pulse obtained by the method of the present invention in the ultra-wideband detection of DC withstand voltage for detecting the discharge of a tip-plate defect in oil and the data diagram during its processing;

[0043] Figure 5 is the original waveform of a typical single PD pulse obtained by the method of the present invention in the ultra-wideband detection of DC withstand voltage for detecting the discharge of an air gap defect inside oil-impregnated paperboard and the data diagram during its processing;

[0044] Figure 6 is the original waveform of a typical single PD pulse obtained by the method of the present invention in the ultra-wideband detection of DC withstand voltage for detecting the discharge of an oil-paper surface defect and the data diagram during its processing;

[0045] Figure 7 is the original waveform of a typical single PD pulse obtained by the method of the present invention in the ultra-wideband detection of DC withstand voltage for detecting the discharge of a suspended defect in oil and the data diagram during its processing.

[0046] Figure 8 is the equivalent time-frequency characteristic plane formed by the method of the present invention, and the parameter distributions corresponding to the pulse waveform-time series (each containing 100 pulse waveforms) of the discharges of tip-plate in oil, air gap defect inside oil-impregnated paperboard, oil-paper surface defect, and suspended defect in oil obtained by the ultra-wideband detection of DC withstand voltage. Detailed Embodiments

[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0048] The present invention provides a method for extracting moment characteristic parameters of partial discharge pulse waveform time-frequency images. For the mixed pulse waveform-time series, i.e., pulse group, formed by PD sources and noise sources and other pulse sources obtained through ultra-wideband detection, first, the single pulse waveform in the pulse group is denoised by wavelet and amplitude-normalized to form a time-domain waveform (or filtering processing may not be performed). The wavelet denoising algorithm is discrete dyadic wavelet transform. Secondly, the time-frequency transform is performed on the time-domain waveform to form a time-frequency image (generalized rectangular time-frequency distribution). Thirdly, the first-order and second-order moments of time and frequency of the time-frequency energy distribution of the time-frequency image are calculated. The sum of the normalized amplitudes under the absolute value of the first-order time moment and the second-order time moment, and the sum of the normalized amplitudes under the absolute value of the first-order frequency moment and the second-order frequency moment are obtained, respectively, to obtain the equivalent time moment and the equivalent frequency moment. Finally, the characteristic parameters are extracted for all pulse waveforms in the pulse group, i.e., the corresponding equivalent time moment and equivalent frequency moment are obtained, and the plane distribution of the equivalent time-frequency moment characteristic parameters is obtained. Based on the plane distribution of the equivalent time-frequency moment characteristic parameters, it can be used to construct a fast classification technology for pulse groups in ultra-wideband PD detection, i.e., for clustering "self-similar" sub-pulse groups to separate multiple PD sources and noise sources.

[0049] This method makes full use of the joint time-frequency distribution characteristics of the non-stationary representation of the pulse waveform signal in the time domain and the frequency domain, and has stronger robustness, simpler algorithm, clearer process, and is easier to implement than the current characteristic parameter extraction methods based on the pulse time domain and frequency domain waveforms respectively (such as Figure 1 the equivalent time-frequency method shown). The most important thing is that it can process pulse sources with similar time-domain characteristics of pulse waveforms and is suitable for constructing a fast classification technology for pulse groups in ultra-wideband PD detection.

[0050] As Figure 2 shown, the system for the method of extracting moment characteristic parameters of partial discharge pulse waveform time-frequency images includes a pulse waveform-time series module (10) for mixing PD sources and noise sources obtained through ultra-wideband detection, a pulse time-domain waveform module (12), a time-frequency transform module (13), a first-order time moment module (14), a second-order time moment module (15), a first-order frequency moment module (16), a second-order frequency moment module (17), an equivalent time moment module (18), an equivalent frequency moment module (19), and an equivalent time-frequency moment characteristic parameter plane distribution module (20);

[0051] The pulse waveform-time series module (10) that mixes the PD source and the noise source obtained by the ultra-wideband detection is connected to the time-frequency transformation module (13) through the pulse time-domain waveform module (12). The time-frequency transformation module (13) is respectively connected to the first-order time moment module (14), the second-order time moment module (15), the first-order frequency moment module (16), and the second-order frequency moment module (17). The first-order time moment module (14) and the second-order time moment module (15) are respectively connected to the equivalent time moment module (18). The first-order frequency moment module (16) and the second-order frequency moment module (17) are respectively connected to the equivalent frequency moment module (19). The equivalent time moment module (18) and the equivalent frequency moment module (19) are respectively connected to the equivalent time-frequency moment characteristic parameter plane distribution module (20).

[0052] The pulse waveform-time series module 10 that mixes the PD source and the noise source obtained by the ultra-wideband detection is a data acquisition device with an analog bandwidth of dozens of MHz and a sampling rate of 100 MS / s or more (a high-pass filter of 10 kHz or more needs to be installed to filter out background noise signals such as high-order harmonics). Through a coupling device with a frequency response of dozens of MHz or more that meets the nanosecond-level PD ultra-wideband detection (which can be used in the PD withstand voltage test circuits under AC, DC, and impulse voltages), based on the pulse waveform triggering technology, the single-pulse time-domain waveform and the pulse waveform-time series corresponding to the triggering moment, that is, the pulse group p j (t), is defined as follows:

[0053]

[0054] In the formula:

[0055] j —— the jth pulse (j = 1, 2,..., N, N is the total number of pulse waveforms included in the pulse group);

[0056] k —— the pulse waveform consists of k points, and the number of points is determined by the sampling rate f s * sampling duration.

[0057] Figure 4 (a), Figure 5 (a), Figure 6 (a) and Figure 7 (a) are examples of single original pulse waveforms included in the pulse group, which are the pulse waveform-time series obtained by the ultra-wideband detection with a sampling rate of 2.5 GS / s and an analog bandwidth of 1 GHz.

[0058] The system also includes a wavelet denoising module 11. The wavelet denoising algorithm is selected as the discrete dyadic wavelet, that is, the single original pulse waveform in the pulse group is denoised by wavelet and amplitude-normalized to form the time-domain waveform, that is, p j (t). For Figure 4 (a),Figure 5 (a), Figure 6 (a) and Figure 7 The waveform after wavelet denoising of the original pulse waveform of PD shown in (a) is as Figure 4 (b), Figure 5 (b), Figure 6 (b) and Figure 7 (b) as shown. The discrete signal p j (t)'s dyadic wavelet, i.e., the discrete dyadic wavelet, is defined as follows:

[0059] ψ(t) ∈ L 2 (R) is a dyadic wavelet, that is

[0060]

[0061] At this time, the wavelet transform of p j (t) at scale 2 is

[0062]

[0063] Then the function sequence

[0064]

[0065] is called the discrete dyadic wavelet transform of p j (t).

[0066] The pulse time-domain waveform module 12 mentioned above is mainly used to store the time-domain waveform, i.e., p j (t), as the object to be processed by subsequent modules. If no wavelet denoising processing is performed, p j (t) is Figure 4 (a), Figure 5 (a) and Figure 6 (a) the single original pulse time-domain waveform shown; if processed by the 11 wavelet denoising module, Figure 4 (b), Figure 5 (b), Figure 6 (b) and Figure 7 (b) the waveforms shown are p j (t), and all are stored in the 12 pulse time-domain waveform module.

[0067] The time-frequency transformation module 13 mentioned above uses the generalized rectangular time-frequency distribution algorithm to perform time-frequency transformation on p j (t) to obtain the time-frequency image tfr j (t, f). The algorithm is defined as follows:

[0068]

[0069] Where: h - rectangular window function.

[0070] Figure 4 (b), Figure 5 (b), Figure 6 (b) and Figure 7 The time-frequency images of the time-domain waveforms after wavelet denoising as shown in (b) are as follows Figure 4 (c), Figure 5 (c), Figure 6 (c) and Figure 7 (c) as shown.

[0071] The first-order time moment module 14, second-order time moment module 15, first-order frequency moment module 16, and second-order frequency moment module 17 are defined as follows in sequence:

[0072]

[0073]

[0074]

[0075]

[0076] Figure 4 (c), Figure 5 (c), Figure 6 (c) and Figure 7 The time-frequency image tfr j (t, f) of the first-order time moment distribution is as follows Figure 4 (d), Figure 5 (d), Figure 6 (d) and Figure 7 (d) as shown; the second-order time moment distribution is as follows Figure 4 (e), Figure 5 (e), Figure 6 (e) and Figure 7 (e) as shown; the first-order frequency moment distribution is as follows Figure 4 (g), Figure 5 (g), Figure 6 (g) and Figure 7 (g) as shown; the second-order frequency moment distribution is as follows Figure 4 (h), Figure 5 (h), Figure 6 (h) and Figure 7 (h) as shown.

[0077] The algorithms of the equivalent time moment module 18 and equivalent frequency moment module 19 are as follows:

[0078]

[0079]

[0080] Figure 4 (c), Figure 5 (c), Figure 6 (c) and Figure 7 The equivalent time moment distribution of the time-frequency graph shown in (c) is as Figure 4 (f), Figure 5 (f), Figure 6 (f) and Figure 7 (f) shown, and the equivalent frequency moment distribution is as Figure 4 (i), Figure 5 (i), Figure 6 (i) and Figure 7 (i) shown. The abscissa values corresponding to the peaks of the equivalent time moment distribution and the equivalent frequency moment distribution are max(T j )_N and max(F j )_N as shown in Table 1 below:

[0081] Table 1

[0082]

[0083] max(T j )_N

[0084]

[0085] The equivalent time-frequency moment characteristic parameter plane distribution module 20 described above is defined as follows:

[0086] TF j =(max(T j )_N, max(F j )_N) (12)

[0087] j - the jth pulse (j = 1, 2,..., N, N is the total number of pulse waveforms included in the pulse group).

[0088] Figure 8 This is the equivalent time-frequency moment characteristic parameter plane distribution formed by the method of the present invention. The direct current withstand voltage ultra-wideband detection obtains the parameter distributions corresponding to the pulse waveforms - time series (each containing 100 pulse waveforms) of the discharges of sharp plates in oil, internal air gaps in oil-impregnated paperboards, surface discharges along oil-paper interfaces, and suspended defects in oil. It can be seen that the single pulse waveforms corresponding to the internal air gaps in oil-impregnated paperboards and the discharges of suspended defects in oil are very similar in the time domain, but Figure 8 The equivalent time-frequency moment characteristic parameter plane distribution formed still shows the characteristics of "intra-class cohesion" and "inter-class separation", which is convenient for later separating the pulse group by means of intelligent clustering analysis and other means to form sub-pulse groups with their respective characteristics, so as to realize the separation of multiple PD sources and noise sources.

[0089] The flowchart of the characteristic parameter processing of the pulse waveform data in the method of the present invention is as Figure 3as shown

[0090] Figures 4 to 8 The figure shows a typical single pulse waveform in the pulse waveform-time series obtained by ultra-wideband detection with a sampling rate of 2.5 GS / s and an analog bandwidth of 1 GHz, in accordance with Figure 3 the single original pulse waveform, the time-domain waveform after wavelet denoising, the time-frequency image of the generalized rectangular time-frequency distribution, the first-order time moment distribution diagram, the second-order time moment distribution diagram, the equivalent time moment distribution diagram, the first-order frequency moment distribution diagram, the second-order frequency moment distribution diagram, and the plane distribution diagram of equivalent time-frequency moment characteristic parameters corresponding to the processing flow as shown

[0091] As described above, the above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims

Claims

1. A method for extracting moment characteristic parameters of the time-frequency image of partial discharge pulse waveforms, characterized in that This method is aimed at obtaining the hybrid pulse waveform-time series, i.e., pulse group, formed by the pulse source for ultra-wideband detection. The method specifically includes the following steps: Step 101: Use wavelet denoising and amplitude normalization for a single pulse waveform in the pulse group to form a time-domain waveform; Step 102: Perform time-frequency transformation on the time-domain waveform to form a time-frequency image; Step 103: Calculate the first-order and second-order moments of time and frequency for the time-frequency energy distribution of the time-frequency image; and calculate the sum of the normalized amplitudes of the absolute values of the first-order time moment and the second-order time moment, and the sum of the normalized amplitudes of the absolute values of the first-order frequency moment and the second-order frequency moment, respectively, to obtain the equivalent time moment and the equivalent frequency moment; Step 104: Extract the characteristic parameters for all pulse waveforms in the pulse group, i.e., obtain the corresponding equivalent time moment and equivalent frequency moment, to get the planar distribution of the equivalent time-frequency moment characteristic parameters; The first-order time moment, the second-order time moment, the first-order frequency moment, and the second-order frequency moment are defined as follows: where f m (t) j is the first-order time moment, B 2 (t) j is the second-order time moment, t m (f) j is the first-order frequency moment, T 2 (f) j is the second-order frequency moment, tfr j (t, f) is the time-frequency image, j is the j-th pulse, where j = 1, 2, … N, and N is the total number of pulse waveforms included in the pulse group; The calculation of the equivalent time moment is as follows: The calculation of the equivalent frequency moment is as follows:

2. A system for the method of extracting moment characteristic parameters of the time-frequency image of partial discharge pulse waveforms described in claim 1, characterized in that, It includes a pulse waveform-time series module (10) for the hybrid of PD source and noise source obtained by ultra-wideband detection, a pulse time-domain waveform module (12), a time-frequency transformation module (13), a first-order time moment module (14), a second-order time moment module (15), a first-order frequency moment module (16), a second-order frequency moment module (17), an equivalent time moment module (18), an equivalent frequency moment module (19), and a planar distribution module (20) of the equivalent time-frequency moment characteristic parameters; The pulse waveform-time series module (10) for the hybrid of PD source and noise source obtained by ultra-wideband detection is connected to the time-frequency transformation module (13) through the pulse time-domain waveform module (12). The time-frequency transformation module (13) is respectively connected to the first-order time moment module (14), the second-order time moment module (15), the first-order frequency moment module (16), and the second-order frequency moment module (17). The first-order time moment module (14) and the second-order time moment module (15) are respectively connected to the equivalent time moment module (18). The first-order frequency moment module (16) and the second-order frequency moment module (17) are respectively connected to the equivalent frequency moment module (19). The equivalent time moment module (18) and the equivalent frequency moment module (19) are respectively connected to the planar distribution module (20) of the equivalent time-frequency moment characteristic parameters.

3. The system according to claim 2, wherein The pulse waveform-time series module (10) for the mixed PD source and noise source obtained by ultra-wideband detection uses a data acquisition device with an analog bandwidth of dozens of MHz and a sampling rate of over 100 MS / s. Through a coupling device with a frequency response of over dozens of MHz to meet the nanosecond-level PD ultra-wideband detection, it records the single-pulse time-domain waveform and the pulse waveform-time series, i.e., the pulse group p j (t), which is defined as follows: In the formula: j is the jth pulse, where j = 1, 2, … N, and N is the total number of pulse waveforms included in the pulse group; The pulse waveform consists of k points, and the number of points is determined by the sampling rate f s * and the sampling duration.

4. The system according to claim 2, wherein The described system further includes a wavelet denoising module (11) disposed between the pulse waveform-time series module (10) and the time-frequency transformation module (13) where the PD source and the noise source are mixed in the ultra-wideband detection and acquisition. The wavelet denoising module (11) is a discrete dyadic wavelet. After wavelet denoising and amplitude normalization of a single original pulse waveform in the pulse group, a time-domain waveform, i.e., p j (t), is formed.

5. The system according to claim 2, characterized in that, The described pulse time-domain waveform module (12) is used to store the time-domain waveform, i.e., p j (t), which serves as the object to be processed by subsequent modules.

6. The system according to claim 2, wherein The time-frequency transformation module (13) performs time-frequency transformation on p j (t) using the generalized rectangular time-frequency distribution algorithm to obtain the time-frequency image tfr j (t, f).

7. The system according to claim 2, wherein The calculation of the planar distribution module (20) of the equivalent time-frequency moment characteristic parameters is as follows: TF j = (max(T j )_N, max(F j )_N) where T j is the equivalent time moment, and F j is the equivalent frequency moment.

Images