UHF Partial Discharge Signal Denoising and Recognition Method Based on Dual-Channel Signal Difference

Abstract

The present invention relates to the technical field of partial discharge detection of power equipment, and discloses a method for denoising and identifying ultra-high frequency partial discharge signals based on dual-channel signal difference, including: S1, using a first ultra-high frequency sensor to collect the initial internal ultra-high frequency signal in the power equipment, and using a second ultra-high frequency sensor to collect the initial spatial ultra-high frequency signal in the space where the power equipment is located; S2, comparing the signal characteristics of the initial internal ultra-high frequency signal and the initial spatial ultra-high frequency signal; S3, respectively performing filtering, envelope detection and amplification processing on the initial internal ultra-high frequency signal and the initial spatial ultra-high frequency signal in sequence; S4, performing feature comparison through peak-time-frequency domain feature analysis method; S8, performing partial discharge analysis according to the PRPS pattern. The present invention can effectively identify whether the detected ultra-high frequency signal is excited by internal partial discharge of power equipment and whether it is interfered by spatial ultra-high frequency electromagnetic waves.

Description

Technical Field

[0001] The present invention relates to the technical field of partial discharge detection of power equipment, and particularly relates to a method for denoising and identifying ultra-high frequency partial discharge signals based on differential of dual-channel signals. Background Art

[0002] When the internal insulation of a power equipment only has partial discharge in a local area without penetrating the entire internal insulation of the equipment, it is called partial discharge (PD) of the power equipment. Partial discharge can occur near the high-voltage conductor or at other positions of the internal insulation. Partial discharge will generate corresponding physical phenomena such as electricity, light, heat, and chemistry. By corroding the insulating medium, it causes further deterioration of the insulation and ultimately leads to equipment failure. Partial discharge can cause the continuous decline of the internal insulation level of the equipment, and it is also a sign of the deterioration of the equipment insulation. Therefore, accurate detection and diagnosis of partial discharge of power equipment help to timely discover and eliminate internal defects of the equipment and ensure the safe and reliable operation of power equipment.

[0003] Every partial discharge inside a power equipment will be accompanied by a steep pulse current. When the rising time of the pulse current reaches the nanosecond level, it can excite ultra-high frequency (UHF) signals with a frequency range of 300 MHz to 3000 MHz. By detecting and analyzing the ultra-high frequency signals excited by partial discharge inside the power equipment, the types, positions, and severity of partial discharge defects inside the power equipment can be discovered and diagnosed, which is the ultra-high frequency partial discharge detection method. Since this method was first applied in the 1980s, it has been proven to have good detection effects on tip discharge, floating potential discharge, and solid insulation discharge inside power equipment. After years of technical research and experience accumulation, it has become increasingly mature. It is an effective and standard means for live detection of partial discharge of power equipment and has been widely used.

[0004] According to a large amount of on-site detection experience, in quite a number of substations and power plants, the electromagnetic environment is complex, and the UHF partial discharge detection method faces relatively large interference. During the detection of partial discharge in power equipment, the detected UHF signals with discharge characteristics may be either the UHF signals excited by internal discharge of the equipment or external interferences including corona discharge of fittings and floating potential discharge, or even the superposition of multiple signals of internal signals and external interferences of the equipment, which causes difficulties in the judgment of the results. In order to distinguish whether the UHF signal comes from inside or outside the equipment, the multi-channel signal time-delay positioning method is mainly used to find the signal source. However, since many power equipment cannot arrange UHF sensors according to the positioning requirements and it is difficult to distinguish co-source signals in the case of multi-source signals, this method is restricted in use in many cases. By means of denoising methods such as software and hardware filtering, wavelet decomposition, and singular value decomposition, UHF interferences in fixed frequency bands such as mobile phone communication signals and radar signals can be filtered out. However, for many external interferences including corona discharge of fittings and floating potential discharge, which themselves also belong to a kind of partial discharge, their time-frequency characteristics are very similar to the UHF signals of internal partial discharge of the equipment, making it difficult to achieve effective denoising and even impossible to analyze the state of power equipment based on the detected signals.

[0005] Therefore, there is an urgent need to design an identification and denoising scheme for UHF partial discharge signals of power equipment in a complex electromagnetic environment to extract effective UHF partial discharge signals for analysis. Summary of the Invention

[0006] The present invention provides a UHF partial discharge signal identification and denoising method based on dual-channel signal difference to solve the problems that the existing multi-channel signal time-delay positioning method, software and hardware filtering method, and denoising algorithm cannot identify and denoise UHF partial discharge signals in a complex electromagnetic environment.

[0007] The present invention is realized through the following technical solutions:

[0008] To solve the above problems, the present application provides a UHF partial discharge signal identification and denoising method based on dual-channel signal difference. This method judges whether the collected UHF partial discharge signal of the power equipment is interfered by spatial UHF electromagnetic waves through the comparison of the signal characteristics of the detection sensor and the background sensor. For the detection signal interfered by spatial UHF electromagnetic waves, a UHF partial discharge signal denoising and identification method based on dual-channel signal difference is used, including:

[0009] S1. Use a first UHF sensor to collect the initial internal UHF signal in the power equipment. At the same time, use a second UHF sensor to collect the initial spatial UHF signal in the space where the power equipment is located;

[0010] S2. Compare the signal characteristics of the initial internal UHF signal and the initial spatial UHF signal to determine whether the internal UHF signal may contain the spatial UHF signal. If so, jump to S3; otherwise, after sequentially performing processes such as detection, amplification, and A / D conversion on the initial internal UHF signal, construct the PRPS spectrum of the initial internal UHF signal and jump to S8;

[0011] S3. Sequentially perform filtering, envelope detection, and amplification processes on the initial internal UHF signal and the initial spatial UHF signal respectively to obtain an intermediate internal UHF signal and an intermediate spatial UHF signal;

[0012] S4. Through the peak-time-frequency domain feature analysis method, compare the characteristics of the intermediate internal UHF signal and the intermediate spatial UHF signal to determine whether the intermediate internal UHF signal contains the intermediate spatial UHF signal. If so, jump to S5; otherwise, directly construct the PRPS spectrum of the intermediate internal UHF signal and jump to S8;

[0013] S5. Align the starting points of the intermediate internal UHF signal and the intermediate spatial UHF signal through the minimum energy method;

[0014] S6. Perform differential processing on the intermediate internal UHF signal and the intermediate spatial UHF signal after starting point alignment to obtain a denoised internal UHF signal;

[0015] S7. After performing A / D conversion on the denoised internal UHF signal, construct the PRPS spectrum of the denoised internal UHF signal;

[0016] S8. Perform partial discharge analysis based on the PRPS spectrum.

[0017] As an optimization, in S1, the second UHF sensor is arranged beside the first UHF sensor, and the detection surface of the second UHF sensor is arranged opposite to the detection surface of the first UHF sensor.

[0018] As an optimization, the specific steps of S2 are as follows:

[0019] The specific steps of S2 are as follows:

[0020] S2.1. On the premise that the first UHF sensor collects the initial internal UHF signal with discharge characteristics, determine whether the second UHF sensor collects the initial spatial UHF signal with discharge characteristics. If so, jump to S2.2; otherwise, after sequentially performing processes such as detection, amplification, and A / D conversion on the initial internal UHF signal, construct the PRPS spectrum of the initial internal UHF signal and jump to S8;

[0021] S2.2. Analyze the pulse signals that make up the initial internal UHF signal and the initial spatial UHF signal respectively according to the signal diagrams of the initial internal UHF signal and the initial spatial UHF signal. If the relative deviation of the pulse duration of the pulse signal that makes up the initial spatial UHF signal from the pulse duration of the pulse signal that makes up the initial internal UHF signal is not greater than 20%, and the maximum relative deviation of the frequency between the pulse signal that makes up the initial spatial UHF signal and the pulse signal that makes up the initial internal UHF signal is not greater than 10%, it is determined that the internal UHF signal may contain the spatial UHF signal.

[0022] As an optimization, the judgment method with discharge characteristics is as follows:

[0023] If the UHF signal has a 50Hz or 100Hz frequency correlation and the single oscillation pulse duration is from ten nanoseconds to one microsecond, it is determined that the UHF signal has discharge characteristics, where the 50Hz or 100Hz frequency correlation means that the oscillation pulse appears 1 or 2 times within one power frequency cycle, and the relative deviation of the time interval of each oscillation pulse is not greater than 5%.

[0024] As an optimization, in S3, the filtering of the initial internal UHF signal and the initial spatial UHF signal is both a band-pass filtering process from 1000MHz to 1200MHz.

[0025] As an optimization, the specific steps of S4 are as follows:

[0026] S4.1. Calculate the peak frequency, frequency centroid, and time-domain centroid of the intermediate internal UHF signal and the intermediate spatial UHF signal respectively;

[0027] The frequency centroid

[0028] The time-domain centroid

[0029] where N is the total number of signal acquisition points, s(t) is the detected signal, F(ω) is the detected signal after Fourier transform, f i is the frequency of the signal at the i-th acquisition point, and t i is the time of the signal at the i-th acquisition point;

[0030] S4.2. Calculate the relative deviation of the peak frequency, the relative deviation of the frequency centroid, and the relative deviation of the time-domain centroid of the intermediate internal UHF signal and the intermediate spatial UHF signal according to the peak frequency, the frequency centroid, and the time-domain centroid of the intermediate internal UHF signal and the intermediate spatial UHF signal;

[0031] S4.3. If the average value of the relative deviation of the peak frequency, the relative deviation of the frequency centroid, and the relative deviation of the time-domain centroid is greater than 30%, it is determined that the intermediate internal UHF signal does not contain the intermediate spatial UHF signal; otherwise, it is determined that the intermediate internal UHF signal contains the intermediate spatial UHF signal.

[0032] As an optimization, the specific steps of S5 are as follows:

[0033] S5.1. Subtract the product of the average energy and the number of acquired points from the cumulative energy to obtain the energy difference curve of the intermediate internal UHF signal and the energy difference curve of the intermediate spatial UHF signal respectively;

[0034] S5.2. Determine the starting points of the intermediate internal UHF signal and the intermediate spatial UHF signal according to the first inflection point from the descent to the ascent of the energy difference curve;

[0035] S5.3. Align the starting points of the intermediate internal UHF signal and the intermediate spatial UHF signal.

[0036] As an optimization, the energy difference curve is expressed by the following formula:

[0037]

[0038] where P represents the average energy of the signal within the entire acquisition time, u i represents the signal voltage value at the i-th acquisition point, N is the total number of acquisition points of the entire signal, and n represents the number of points that have been acquired.

[0039] As an optimization, the specific steps of S6 are as follows:

[0040] S6.1. At each acquisition point, subtract the intermediate spatial UHF from the aligned intermediate internal UHF signal at the starting point to obtain the differential calculation result at each acquisition point;

[0041] S6.2. If the differential calculation result is positive, retain the differential calculation result; if the differential calculation result is an amplitude value, set the differential calculation result to zero. All the differential calculation results that are finally positive form the denoised internal UHF signal.

[0042] As an optimization, the specific steps of S7 are as follows:

[0043] S7.1. Divide the pulse sequence unit obtained after A / D conversion of the denoised internal UHF signal equally according to the phase to obtain the distribution of each phase pulse;

[0044] S7.2. Arrange the pulse sequence units obtained after A / D conversion of the denoised internal UHF signals according to the occurrence times within 50 power frequency cycles to obtain the pulse distributions of each time series.

[0045] S7.3. Draw the PRPS map of the UHF signals inside the power equipment after denoising based on the analysis results of the phase pulse distribution and the corresponding time series pulse distribution of the pulse sequence units.

[0046] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0047] The present invention overcomes the problems that the existing positioning methods, filtering methods, and denoising algorithms cannot effectively identify and denoise the UHF partial discharge signals inside the substation equipment under complex electromagnetic environments, especially when there is continuous corona discharge or floating potential discharge of fittings or even multi-source signals outside. It can effectively identify whether the detected UHF signal is excited by internal partial discharge of power equipment and whether it is interfered by spatial UHF electromagnetic waves with a relatively simple operation process, calculation process, and relatively low-cost hardware device; for the detected signals interfered by spatial UHF electromagnetic waves, it can effectively remove the interference. Although some discharge information may be lost to a certain extent, the denoised UHF partial discharge signals can still be used for the analysis and diagnosis of partial discharge of substation equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts. In the drawings:

[0049] Figure 1 It is a flowchart of a method for identifying and denoising UHF partial discharge signals based on differential of dual-channel signals according to the present invention.

[0050] FIG. 2(a) is a schematic diagram of the UHF signal inside the power equipment collected on-site by the first UHF sensor provided in the embodiment of the present invention through an oscilloscope.

[0051] FIG. 2(b) is a schematic diagram of the spatial background UHF signal collected on-site by the second UHF sensor provided in the embodiment of the present invention through an oscilloscope.

[0052] Figure 3 It is a schematic diagram of the PRPS map of the UHF signal inside the power equipment after denoising provided in the embodiment of the present invention.

[0053] Figure 4Schematic diagram of the foreign object discharge defect of the gas-insulated metal-enclosed switchgear housing corresponding to the phase-resolved pulse sequence (PRPS) map of the ultra-high frequency signal inside the power equipment after denoising provided by the embodiments of the present invention. Detailed implementation manners

[0054] To make the objectives, technical solutions and advantages of the present invention more clear and understandable, the present invention will be further described in detail below in conjunction with embodiments and drawings. The illustrative embodiments and descriptions thereof of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0055] Before the description of the embodiments, the following terms are explained first:

[0056] PRPS: Phase Resolved Pulse Sequence, the phase distribution of the pulse sequence. The X-axis is the phase, showing the equally divided phases of a 20-millisecond power frequency cycle; the Y-axis is the time, showing 50 power frequency cycles; the Z-axis is the amplitude of the discharge pulse.

[0057] A method for denoising and identifying ultra-high frequency partial discharge signals based on dual-channel signal difference in Embodiment 1, as Figure 1 shown, includes:

[0058] S1. Use the first ultra-high frequency sensor to collect the initial internal ultra-high frequency signal inside the power equipment. At the same time, use the second ultra-high frequency sensor to collect the initial spatial ultra-high frequency signal in the space where the power equipment is located;

[0059] In S1, press the detection surface of the first ultra-high frequency sensor tightly against the metal discontinuities such as basin insulators, observation windows, and flange gaps to collect the initial internal ultra-high frequency signal of the power equipment.

[0060] The second ultra-high frequency sensor is arranged beside the first ultra-high frequency sensor, and the detection surface of the second ultra-high frequency sensor is arranged opposite to the detection surface of the first ultra-high frequency sensor. Facing away from the power equipment can better detect the ultra-high frequency electromagnetic wave interference in the space, and doing so can separate the internal partial discharge signal and the space interference signal for detection as much as possible.

[0061] Place the second ultra-high frequency sensor back-to-back with the first ultra-high frequency sensor, or place the ultra-high frequency sensor adjacent to the first ultra-high frequency sensor and the detection surface facing away from the power equipment to collect the spatial ultra-high frequency (electromagnetic wave) signal in the adjacent space, that is, the background electromagnetic wave signal.

[0062] S2. Compare the signal characteristics of the initial internal UHF signal and the initial spatial UHF signal to determine whether the internal UHF signal may contain spatial UHF signals. If so, jump to S3; otherwise, after processing the initial internal UHF signal through detection, amplification, A / D conversion, etc., construct the PRPS spectrum of the initial internal UHF signal and jump to S8.

[0063] Determine whether the original internal UHF signal of the power equipment collected by the first UHF sensor is interfered by spatial UHF electromagnetic waves.

[0064] The specific steps are as follows:

[0065] S2.1. On the premise that the first UHF sensor collects the initial internal UHF signal with discharge characteristics, determine whether the second UHF sensor collects the initial spatial UHF signal with discharge characteristics. If so, jump to S2.2; otherwise, after processing the initial internal UHF signal through detection, amplification, A / D conversion, etc., construct the PRPS spectrum of the initial internal UHF signal and jump to S8.

[0066] That is, when the first UHF sensor collects an abnormal UHF signal with discharge characteristics and the second UHF sensor does not collect an abnormal UHF signal, it is determined that the initial internal UHF signal of the power equipment collected by the first UHF sensor is not interfered by the initial spatial UHF electromagnetic wave.

[0067] When the initial internal UHF signal of the power equipment collected by the first UHF sensor is not interfered by the initial spatial UHF electromagnetic wave, it can be directly processed through detection, amplification, A / D conversion, etc., and the PRPS spectrum of the initial internal UHF signal of the power equipment can be constructed for partial discharge analysis.

[0068] S2.2. Analyze the pulse signals that make up the initial internal UHF signal and the initial spatial UHF signal respectively according to the signal schematic diagrams of the initial internal UHF signal and the initial spatial UHF signal. If the relative deviation of the pulse duration of the pulse signal that makes up the initial spatial UHF signal from the pulse duration of the pulse signal that makes up the initial internal UHF signal is not more than 20%, and the maximum frequency relative deviation between the pulse signal that makes up the initial spatial UHF signal and the pulse signal that makes up the initial internal UHF signal is not more than 10%, it is determined that the internal UHF signal may contain spatial UHF signals.

[0069] When both the first UHF sensor and the second UHF sensor collect abnormal UHF signals with discharge characteristics, it is determined that the internal UHF signal of the power equipment collected by the first UHF sensor may be interfered by spatial UHF electromagnetic waves.

[0070] In this embodiment, the judgment method with discharge characteristics is as follows:

[0071] If the UHF signal has a 50 Hz or 100 Hz frequency correlation and the single oscillation pulse lasts from ten nanoseconds to one microsecond, it is determined that the UHF signal has discharge characteristics, where the 50 Hz or 100 Hz frequency correlation means that the oscillation pulse appears once or twice within one power frequency cycle, and the relative deviation of the time interval of each oscillation pulse is not greater than 5%.

[0072] (a) in Figure 2 is the initial internal UHF signal of the power equipment collected on-site by the first UHF sensor through an oscilloscope. This signal shows an obvious 100 Hz frequency correlation. However, in addition to the pulses with relatively short duration and relatively large amplitude, there are also pulses with relatively long duration and relatively small amplitude, indicating that the collected signal contains multiple source signals.

[0073] (b) in Figure 2 is the spatial UHF signal collected on-site by the second UHF sensor through an oscilloscope. This signal shows an obvious 100 Hz frequency correlation. Many pulses are similar to the signal in (a) of Figure 2 and have a larger amplitude than the signal in (a) of Figure 2, but the signal types contained are not as rich as the signal in (a) of Figure 2.

[0074] The signals in (a) and (b) of Figure 2 both have obvious discharge characteristics and some of their pulses are similar. It is determined that the initial internal UHF signal of the power equipment collected by the first UHF sensor may be interfered by the spatial UHF electromagnetic wave.

[0075] S3. Filter, envelope detect, and amplify the initial internal UHF signal and the initial spatial UHF signal in sequence to obtain an intermediate internal UHF signal and an intermediate spatial UHF signal.

[0076] When the initial internal original UHF signal of the power equipment collected by the first UHF sensor may be interfered by the initial spatial UHF electromagnetic wave, the filtering of the initial internal UHF signal and the initial spatial UHF signal is both a band-pass filtering process of 1000 MHz to 1200 MHz.

[0077] That is, when the initial internal UHF signal of the power equipment collected by the first UHF sensor may be interfered by the initial spatial UHF electromagnetic wave, perform a hardware band-pass filtering process of 1000 MHz to 1200 MHz on the signal collected by the first UHF sensor.

[0078] Meanwhile, in the case where the initial internal UHF signal of the power equipment collected by the first UHF sensor may be interfered by the initial spatial UHF electromagnetic wave, the signal collected by the second UHF sensor is subjected to a hardware band-pass filtering process from 1000 MHz to 1200 MHz.

[0079] Performing a hardware band-pass filtering process on the signal from 1000 MHz to 1200 MHz can first filter out the UHF electromagnetic wave interference in the low-frequency band existing in the space, laying a foundation for the subsequent signal recognition and denoising.

[0080] Then, the signal collected by the first UHF sensor after filtering is subjected to envelope detection and amplification processing to obtain an intermediate internal UHF signal, so that the frequency of the intermediate internal UHF signal after detection does not exceed 50 MHz. The band-pass filtering from 1000 MHz to 1200 MHz is to try whether the spatial UHF electromagnetic wave interference can be directly filtered out to obtain a narrow-band UHF partial discharge signal. And the subsequent detection reduces the frequency of the narrow-band UHF signal so that it does not exceed 50 MHz, which can basically retain the narrow-band signal profile. The purpose is to enable signal processing with a low sampling rate device, reducing the volume and cost of the device.

[0081] Meanwhile, the signal collected by the second UHF sensor after filtering is also subjected to envelope detection and amplification processing to obtain an intermediate spatial UHF signal, so that the frequency of the intermediate spatial UHF electromagnetic wave after detection does not exceed 50 MHz.

[0082] The basic circuit of envelope detection consists of a diode and an R L C low-pass filter connected in series. During the detection process, the average voltage output at both ends of the R L reflects the envelope change law. The larger the time constant R L C, the better the detection performance. However, considering the limitations of the calculation process and the hardware device, the time constant R L C should be selected while taking into account both the detection performance and the requirement that the frequency of the signal after detection does not exceed 50 MHz.

[0083] S4. By using the peak-time-frequency domain feature analysis method, the intermediate internal UHF signal and the intermediate spatial UHF signal are subjected to feature comparison to determine whether the intermediate internal UHF signal contains the intermediate spatial UHF signal. If so, jump to S5; otherwise, directly construct the PRPS spectrum of the intermediate internal UHF signal and jump to S8.

[0084] By using the peak-time-frequency domain feature analysis method, the signals collected by the first ultra-high frequency sensor and the second ultra-high frequency sensor after detection are compared in terms of features, and determining whether the intermediate internal ultra-high frequency signal is interfered by the intermediate spatial ultra-high frequency electromagnetic wave can be regarded as determining whether the initial internal ultra-high frequency signal of the power equipment collected by the first ultra-high frequency sensor is interfered by the initial spatial ultra-high frequency electromagnetic wave. The peak-time-frequency domain features include the peak frequency f p 、the frequency centroid W, and the time domain centroid t o .

[0085] The peak frequency f p is the frequency corresponding to the part with the largest amplitude in the collected signal.

[0086] The specific steps are as follows:

[0087] S4.1. Calculate the peak frequency, frequency centroid, and time domain centroid of the intermediate internal ultra-high frequency signal and the intermediate spatial ultra-high frequency signal respectively;

[0088] That is, calculate the peak frequency f p 、the frequency centroid W, and the time domain centroid t o of the signal collected by the first ultra-high frequency sensor after detection.

[0089] Calculate the peak frequency f p 、the frequency centroid W, and the time domain centroid t o of the signal collected by the second ultra-high frequency sensor after detection.

[0090] The calculation formula for the frequency centroid is:

[0091] The calculation formula for the time domain centroid is:

[0092] where N is the total number of signal acquisition points, s(t) is the detected signal, F(ω) is the detected signal after Fourier transform, f i is the frequency of the signal at the i-th acquisition point, and t i is the time of the signal at the i-th acquisition point;

[0093] S4.2. Calculate the relative deviation of the peak frequency, the relative deviation of the frequency centroid, and the relative deviation of the time domain centroid of the intermediate internal ultra-high frequency signal and the intermediate spatial ultra-high frequency signal according to the peak frequency, frequency centroid, and time domain centroid of the intermediate internal ultra-high frequency signal and the intermediate spatial ultra-high frequency signal;

[0094] Relative deviation = |(intermediate internal ultra-high frequency signal - intermediate spatial background ultra-high frequency signal) / intermediate internal ultra-high frequency signal|.

[0095] Calculate the peak frequency f of the signals collected by the first UHF sensor and the second UHF sensor after detection p of the relative deviation, the relative deviation of the frequency centroid W of the signals collected by the first UHF sensor and the second UHF sensor after detection, and the time-domain centroid t of the signals collected by the first UHF sensor and the second UHF sensor after detection o relative deviation.

[0096] S4.3. If the average value of the relative deviation of the peak frequency, the relative deviation of the frequency centroid, and the relative deviation of the time-domain centroid is greater than 30%, it is determined that the intermediate internal UHF signal does not contain the intermediate spatial UHF signal; otherwise, it is determined that the intermediate internal UHF signal contains the intermediate spatial UHF signal.

[0097] If the average relative deviation of the three characteristic values is greater than 30%, it is determined that the intermediate internal UHF signal of the power equipment collected by the first UHF sensor is not interfered by the intermediate spatial UHF electromagnetic wave. If the average relative deviation of the three characteristic values is less than or equal to 30%, it is determined that the intermediate internal UHF signal of the power equipment collected by the first UHF sensor is interfered by the intermediate spatial UHF electromagnetic wave.

[0098] In the case that the internal UHF signal of the power equipment collected by the first UHF sensor after filtering and detection is not interfered by the spatial UHF electromagnetic wave, direct A / D conversion processing can be carried out to construct the PRPS spectrum of the intermediate internal UHF signal of the power equipment for partial discharge analysis.

[0099] S5. Align the starting points of the intermediate internal UHF signal and the intermediate spatial UHF signal by the minimum energy method;

[0100] In the case that the internal UHF signal of the power equipment collected by the first UHF sensor is interfered by the spatial UHF electromagnetic wave, align the starting points of the signals collected by the first UHF sensor and the second UHF sensor after detection by the minimum energy method. The minimum energy method refers to subtracting the average energy from the cumulative energy to obtain the energy difference curve of the signal to be analyzed, and this curve can reflect the signal mutation and the overall trend.

[0101] The specific steps are as follows:

[0102] S5.1. Subtract the product of the average energy and the number of points already collected from the cumulative energy to obtain the energy difference curve of the intermediate internal UHF signal and the energy difference curve of the intermediate spatial UHF signal respectively;

[0103] Calculate the energy difference curve of the signal collected by the first UHF sensor after detection, and calculate the energy difference curve of the signal collected by the second UHF sensor after detection.

[0104] In this embodiment, the energy difference curve is expressed by the following formula:

[0105]

[0106] where P represents the average signal energy during the entire acquisition time, u i represents the signal voltage value at the i-th acquisition point (which can be understood as the amplitude of the acquisition point in Figure 2), N is the total number of acquisition points of the entire signal, and n represents the number of points that have been acquired.

[0107] S5.2. Determine the starting points of the intermediate internal UHF signal and the intermediate spatial UHF signal according to the inflection point where the energy difference curve drops to rises;

[0108] That is, according to the inflection point where the energy difference curve drops to rises, determine the starting point of the acquisition signal of the first UHF sensor after detection; according to the inflection point where the energy difference curve drops to rises, determine the starting point of the acquisition signal of the second UHF sensor after detection.

[0109] S5.3. Align the starting points of the intermediate internal UHF signal and the intermediate spatial UHF signal. Alignment means that the starting points of the two signals are aligned on the time axis.

[0110] S6. Perform differential processing on the intermediate internal UHF signal and the intermediate spatial UHF signal after aligning the starting points to obtain the denoised internal UHF signal;

[0111] The specific steps are as follows:

[0112] S6.1. At each acquisition point, subtract the intermediate spatial UHF from the intermediate internal UHF signal after aligning the starting points to obtain the differential calculation result at each acquisition point;

[0113] S6.2. If the differential calculation result is positive, retain the differential calculation result. If the differential calculation result is an amplitude value, set the differential calculation result to zero. All the differential calculation results that are finally positive form the denoised internal UHF signal.

[0114] For the differential calculation results at each acquisition point, positive numbers are retained and negative numbers are set to zero.

[0115] Differential calculation results at each acquisition point

[0116] In the formula, f(x, y) is the acquisition signal of the first UHF sensor at the corresponding phase, that is, the detection signal, and B(x, y) is the acquisition signal of the second UHF sensor at the corresponding phase, that is, the spatial signal.

[0117] Obtain the waveform of the denoised internal UHF signal.

[0118] S7. Convert the denoised internal UHF signal into a digital signal through A / D conversion, and then construct a PRPS spectrum of the denoised internal UHF signal;

[0119] The specific steps are as follows:

[0120] S7.1. Divide the pulse sequence units obtained after A / D conversion of the denoised internal UHF signal at equal phase intervals to obtain the pulse distributions at each phase;

[0121] A / D conversion includes four processes: sampling, holding, quantization, and encoding. A 16-bit resolution A / D converter is used, and the sampling rate f s = 1 / T is at least 100 MHz to ensure signal integrity.

[0122] S7.2. Arrange the pulse sequence units obtained after A / D conversion of the denoised internal UHF signal according to the time of occurrence within 50 power frequency cycles to obtain the pulse distributions of each time series;

[0123] S7.3. Draw a PRPS spectrum of the UHF signal inside the power equipment after denoising based on the analysis results of the phase pulse distribution and the corresponding time series pulse distribution of the pulse sequence units.

[0124] Figure 3 It is a schematic diagram of the PRPS spectrum of the UHF signal inside the power equipment after denoising. This signal has a significant 50 Hz frequency correlation, with a low signal amplitude, a large phase width of the pulse, and shows the characteristics of a tip discharge signal. Except for occasional pulses that do not affect the judgment, all interference signals have been filtered out.

[0125] S8. Conduct partial discharge analysis based on the PRPS spectrum.

[0126] Figure 4 This is based on Figure 3 The signal diagnosis and analysis found a foreign object discharge defect in the shell of the gas-insulated metal-enclosed switchgear. The bolt was left beside the support insulator during the infrastructure installation, and tip discharge occurred during the operation of the equipment.

[0127] The above specific implementation manners further elaborate the purpose, technical solutions, and beneficial effects of the present invention. It should be understood that the above are only specific implementation manners of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for denoising and identifying ultra-high frequency partial discharge signals based on differential of dual-channel signals, characterized in that, it includes: S1. Use a first ultra-high frequency sensor to collect the initial internal ultra-high frequency signal in the power equipment. At the same time, use a second ultra-high frequency sensor to collect the initial spatial ultra-high frequency signal in the space where the power equipment is located; S2. Compare the signal characteristics of the initial internal ultra-high frequency signal and the initial spatial ultra-high frequency signal to determine whether the internal ultra-high frequency signal may contain the spatial ultra-high frequency signal. If so, jump to S3. Otherwise, after sequentially performing detection, amplification, and A / D conversion on the initial internal ultra-high frequency signal, construct the PRPS map of the initial internal ultra-high frequency signal, and jump to S8; S3. Sequentially perform filtering, envelope detection, and amplification on the initial internal ultra-high frequency signal and the initial spatial ultra-high frequency signal respectively to obtain an intermediate internal ultra-high frequency signal and an intermediate spatial ultra-high frequency signal; S4. Through the peak-time-frequency domain feature analysis method, compare the characteristics of the intermediate internal ultra-high frequency signal and the intermediate spatial ultra-high frequency signal to determine whether the intermediate internal ultra-high frequency signal contains the intermediate spatial ultra-high frequency signal. If so, jump to S5. Otherwise, directly construct the PRPS map of the intermediate internal ultra-high frequency signal and jump to S8; S5. Align the starting points of the intermediate internal ultra-high frequency signal and the intermediate spatial ultra-high frequency signal by the minimum energy method; S6. Perform differential processing on the intermediate internal ultra-high frequency signal and the intermediate spatial ultra-high frequency signal after starting point alignment to obtain a denoised internal ultra-high frequency signal; S7. After performing A / D conversion on the denoised internal ultra-high frequency signal, construct the PRPS map of the denoised internal ultra-high frequency signal; S8. Construct the PRPS map of the corresponding internal ultra-high frequency signal and perform partial discharge analysis based on the PRPS map.

2. The method for denoising and identifying ultra-high frequency partial discharge signals based on differential of dual-channel signals according to claim 1, characterized in that, in S1, the second ultra-high frequency sensor is arranged beside the first ultra-high frequency sensor, and the detection surface of the second ultra-high frequency sensor is arranged opposite to the detection surface of the first ultra-high frequency sensor.

3. The method for denoising and identifying ultra-high frequency partial discharge signals based on differential of dual-channel signals according to claim 1, characterized in that, the specific steps of S2 are: the specific steps of S2 are: S2.

1. On the premise that the first ultra-high frequency sensor collects the initial internal ultra-high frequency signal with discharge characteristics, determine whether the second ultra-high frequency sensor collects the initial spatial ultra-high frequency signal with discharge characteristics. If so, jump to S2.

2. Otherwise, after sequentially performing detection, amplification, A / D conversion, etc. on the initial internal ultra-high frequency signal, construct the PRPS map of the initial internal ultra-high frequency signal, and jump to S8; S2.

2. Analyze the pulse signals that make up the initial internal UHF signal and the initial spatial UHF signal respectively according to the signal schematic diagrams of the initial internal UHF signal and the initial spatial UHF signal. If the relative deviation of the pulse duration of the pulse signals that make up the initial spatial UHF signal from the pulse duration of the pulse signals that make up the initial internal UHF signal is no more than 20%, and the maximum relative frequency deviation between the pulse signals that make up the initial spatial UHF signal and the pulse signals that make up the initial internal UHF signal is no more than 10%, then it is determined that the internal UHF signal may contain the spatial UHF signal.

4. A method for denoising and identifying UHF partial discharge signals based on dual-channel signal difference according to claim 3, characterized in that, the judgment method with discharge characteristics is: If the UHF signal has a 50 Hz or 100 Hz frequency correlation and the single oscillation pulse lasts from ten nanoseconds to one microsecond, then it is judged that the UHF signal has discharge characteristics, where the 50 Hz or 100 Hz frequency correlation means that the oscillation pulse appears 1 or 2 times within one power frequency cycle, and the relative deviation of the time interval of each oscillation pulse is no more than 5%.

5. A method for denoising and identifying UHF partial discharge signals based on dual-channel signal difference according to claim 1, characterized in that, In S3, the filtering of the initial internal UHF signal and the initial spatial UHF signal is both band-pass filtering processing from 1000 MHz to 1200 MHz.

6. A method for denoising and identifying UHF partial discharge signals based on dual-channel signal difference according to claim 1, characterized in that, The specific steps of S4 are: S4.

1. Calculate the peak frequency, frequency centroid and time-domain centroid of the intermediate internal UHF signal and the intermediate spatial UHF signal respectively; The frequency centroid The time-domain centroid Where N is the total number of signal acquisition points, s(t) is the detected signal, F(ω) is the detected signal after Fourier transform, and f i is the frequency of the signal at the i-th acquisition point, and t i is the time of the signal at the i-th acquisition point; S4.

2. Calculate the relative deviation of the peak frequency, the relative deviation of the frequency centroid and the relative deviation of the time-domain centroid of the intermediate internal UHF signal and the intermediate spatial UHF signal according to the peak frequency, frequency centroid and time-domain centroid of the intermediate internal UHF signal and the intermediate spatial UHF signal; S4.

3. If the average value of the relative deviation of the peak frequency, the relative deviation of the frequency centroid and the relative deviation of the time-domain centroid is greater than 30%, then it is determined that the intermediate internal UHF signal does not contain the intermediate spatial UHF signal, otherwise, it is determined that the intermediate internal UHF signal contains the intermediate spatial UHF signal.

7. A method for denoising and identifying UHF partial discharge signals based on dual-channel signal difference according to claim 6, characterized in that, The specific steps of S5 are: S5.

1. Subtract the product of the average energy and the number of acquired points from the cumulative energy to obtain the energy difference curve of the intermediate internal UHF signal and the energy difference curve of the intermediate spatial UHF signal respectively; S5.

2. Determine the starting points of the intermediate internal UHF signal and the intermediate spatial UHF signal according to the first inflection point from the decline to the rise of the energy difference curve; S5.

3. Align the starting points of the intermediate internal UHF signal and the intermediate spatial UHF signal.

8. A method for denoising and identifying ultra-high frequency partial discharge signals based on dual-channel signal difference according to claim 7, characterized in that, the energy difference curve is expressed by the following formula: Where P represents the average energy of the signal during the entire acquisition time, u i represents the signal voltage value at the i-th acquisition point, N is the total number of acquisition points of the entire signal, and n represents the number of points that have been acquired.

9. A method for denoising and identifying ultra-high frequency partial discharge signals based on dual-channel signal difference according to claim 1, characterized in that, the specific steps of S6 are: S6.

1. At each acquisition point, subtract the middle spatial ultra-high frequency signal from the aligned middle internal ultra-high frequency signal at the starting point to obtain the differential calculation result at each acquisition point; S6.

2. If the differential calculation result is positive, retain the differential calculation result. If the differential calculation result is an amplitude value, set the differential calculation result to zero. All the differential calculation results that are finally positive form the denoised internal ultra-high frequency signal.

10. A method for denoising and identifying ultra-high frequency partial discharge signals based on dual-channel signal difference according to claim 1, characterized in that, the specific steps of S7 are: S7.

1. Divide the pulse sequence unit obtained after A / D conversion of the denoised internal ultra-high frequency signal at equal distances according to the phase to obtain the pulse distribution of each phase; S7.

2. Arrange the pulse sequence unit obtained after A / D conversion of the denoised internal ultra-high frequency signal according to the time of occurrence within 50 power frequency cycles to obtain the pulse distribution of each time series; S7.

3. According to the analysis results of the phase pulse distribution and the corresponding time series pulse distribution of the pulse sequence unit, draw the PRPS map of the ultra-high frequency signal inside the power equipment after denoising.

Images