Method for identifying main and longitudinal insulation defects of converter transformer valve side under complex electric stress

Abstract

The invention discloses a method for identifying main longitudinal insulation defects of a converter transformer valve side under complex electrical stress, which comprises the following steps of: 1, acquiring a partial discharge signal of the converter transformer valve side, recording the amplitude and phase of each partial discharge by taking T as a period, and drawing a PRPD spectrogram when the number of discharge times is accumulated to exceed N; 2, describing a PRPD spectrogram profile to obtain a PRPD spectrogram profile curve so as to represent the change condition of the maximum discharge amplitude along with the phase; 3, performing frequency domain analysis on the PRPD spectrogram profile curve; and 4, identifying main insulation discharge and longitudinal insulation discharge. According to the method, under the condition that the complex electric stress effect of power frequency alternating current, direct current and high-proportion harmonic waves on the valve side of the converter transformer is considered, the PRPD spectrogram characteristics are extracted, main insulation and longitudinal insulation discharge distinguishing is achieved, and targeted measures can be conveniently taken for different insulation defects in actual operation.

Description

Technical Field

[0001] The present invention belongs to the field of high voltage and insulation technology, and specifically relates to a method for identifying main longitudinal insulation defects under complex electrical stress on the valve side of a converter transformer. Background Art

[0002] The converter transformer is the core component of the high-voltage direct current (HVDC) transmission system, and undertakes the important tasks of voltage conversion and isolation between AC and DC systems. The valve-side winding insulation, as the weak link of the converter transformer, is one of the main factors leading to converter transformer accidents. The oil-paper insulation on the valve side of the converter transformer has been subjected to the combined effects of various voltage stresses such as power frequency AC, DC, and high-proportion harmonics for a long time, resulting in more complex local discharge characteristics and mechanisms, more prominent insulation problems, and more difficult insulation status assessment. Compared with traditional power transformers, the rapidly changing harmonic components in the converter transformer increase the steepness of the voltage waveform, which not only accelerates the bombardment and erosion of the charge on the oil-paper insulation, but also highlights the hysteresis of the interface charge migration with voltage changes, thereby causing more severe local discharge. Based on the extraction and analysis of various local discharge feature quantities, domestic and foreign scholars have applied artificial intelligence methods such as machine learning and deep neural networks to construct a series of reliable pattern recognition methods.

[0003] However, existing research mainly faces the following difficulties: First, existing pattern recognition methods focus on identifying insulation defect types, and fail to distinguish whether defects occur in main insulation discharge or longitudinal insulation discharge. Therefore, it is difficult to make accurate countermeasures for insulation defects in different locations in field applications. Second, the research on existing pattern recognition methods mainly focuses on partial discharges under DC, industrial frequency AC and AC / DC composite voltages, which are somewhat different from the complex electrical stresses with a high proportion of harmonics superimposed in converter transformers. Therefore, it is difficult to apply to the insulation defect discharge scenario on the valve side of the converter transformer. In the actual operation of the valve side of the converter transformer, different treatment methods are usually adopted for insulation defects occurring in different locations. Therefore, it is crucial to identify whether the insulation defects occur in the main insulation or the longitudinal insulation. Summary of the invention

[0004] In order to solve the problems existing in the above-mentioned prior art, the purpose of the present invention is to provide a method for identifying main longitudinal insulation defects under complex electrical stress on the valve side of a converter transformer. The method of the present invention realizes the distinction between main insulation and longitudinal insulation discharge by extracting PRPD spectrum features under the action of complex electrical stress of industrial frequency AC, DC, and high-proportion harmonics on the valve side of the converter transformer, which is convenient for taking targeted measures for different insulation defects in actual operation.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A method for identifying main longitudinal insulation defects under complex electrical stress on the valve side of a converter transformer comprises the following steps:

[0007] Step 1: Collect partial discharge information and make PRPD spectrum;

[0008] Collect the partial discharge signal on the valve side of the converter transformer, record the amplitude and phase of each partial discharge with a period of T, and draw a PRPD spectrum when the cumulative number of discharges exceeds N;

[0009] Step 2: Describe the PRPD spectrum profile and obtain the PRPD spectrum profile curve to show the change of the maximum discharge amplitude with the phase;

[0010] Step 3: Perform frequency domain analysis on the PRPD spectrum profile curve;

[0011] Step 4: Identify main insulation and longitudinal insulation discharge

[0012] 1) If the amplitude of the 900Hz component accounts for the largest percentage of the fundamental frequency, it is longitudinal insulation discharge;

[0013] 2) If the 900Hz component is not the largest and the 100Hz component is greater than the 300Hz component, it is longitudinal insulation discharge;

[0014] 3) If the 900Hz component is not the largest and the 100Hz component is less than the 300Hz component, the main insulation discharge occurs.

[0015] Preferably, in the first step, the acquisition method for collecting the local discharge signal on the valve side of the converter transformer adopts a pulse current method, an ultra-high frequency method or an ultrasonic method. The pulse current method has the advantages of high sensitivity and strong anti-interference ability; the ultra-high frequency method has good anti-corona performance and is suitable for online monitoring of local discharge; the ultrasonic method is simple to operate on site and convenient to use; the acquisition form adopts continuous acquisition or triggered acquisition. Continuous acquisition can obtain all subtle discharge signals and improve the sensitivity of identification; triggered acquisition can filter out part of the signal interference caused by environmental reasons and improve the reliability of identification.

[0016] Preferably, in the first step, T is an integer multiple of 20ms; N≥100; T is an integer multiple of 20ms, that is, an integer multiple of the power frequency period, which is convenient for PRPD spectrum contour characterization of discharge signals within one or more cycles; N≥100, that is, when the discharge points reach more than 100, judgment and identification are performed, in order to reduce the impact of partial discharge randomness and improve the reliability of identification results.

[0017] Preferably, the method for depicting the PRPD spectrum profile in the second step is: the PRPD spectrum is longitudinally cut into different phase regions at intervals of D, the maximum discharge amplitude in each phase region is extracted, and the maximum discharge amplitudes of each phase region are sequentially connected to obtain the PRPD spectrum profile curve. This method is simple and convenient, and can complete the depiction of the PRPD spectrum profile curve in a relatively short time, thereby improving the speed of identification.

[0018] Preferably, D is less than or equal to 5°. When D is less than or equal to 5°, the cutting of the PRPD spectrum is fine enough and the characterization of the spectrum contour is reliable enough. The smaller D is, the more accurate the characterization of the spectrum contour is and the more reliable the identification result is.

[0019] Preferably, the frequency domain analysis in the third step adopts fast Fourier analysis. Compared with the traditional discrete Fourier transform, fast Fourier analysis greatly reduces the amount of calculation, which makes fast Fourier analysis more efficient when processing large amounts of data, especially in real-time signal processing.

[0020] Compared with the prior art, the present invention has the following advantages:

[0021] Taking into account the complex electrical stress of power frequency AC, DC, and high-proportion harmonics on the valve side of the converter transformer, the main insulation and longitudinal insulation discharge are distinguished by extracting the PRPD spectrum features, which facilitates the adoption of targeted measures for different insulation defects in actual operation. Compared with the prior art, the present invention takes into account the complex electrical stress on the valve side of the converter transformer, so it is more suitable for the valve side of the converter transformer, and when a partial discharge signal appears on site, detection can be started immediately, and identification results can be given at the early stage of partial discharge development, which can leave enough reaction time for on-site staff to deal with the faults that occur. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 is a flow chart of the method of the present invention.

[0023] Figure 2 It is the PRPD spectrum of pure oil gap insulation with pin-plate defect of main insulation, where (a) is when the discharge times are accumulated to 100 times, (b) is when the discharge times are accumulated to 500 times, and (c) is when the discharge times are accumulated to 1000 times.

[0024] Figure 3 It is the contour curve of the PRPD spectrum of pure oil gap insulation with needle plate defect of the main insulation, where (a) is the cumulative discharge number of 100 times, (b) is the cumulative discharge number of 500 times, and (c) is the cumulative discharge number of 1000 times.

[0025] Figure 4 It is the frequency domain analysis result of the main insulation PRPD spectrum contour curve, where (a) is the cumulative discharge number of 100 times, (b) is the cumulative discharge number of 500 times, and (c) is the cumulative discharge number of 1000 times. DETAILED DESCRIPTION

[0026] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0027] The following is an implementation case using the pin plate defect pure oil gap insulation of the main insulation as an example.

[0028] like Figure 1 As shown, the present invention provides a method for identifying main longitudinal insulation defects under complex electrical stress on the valve side of a converter transformer, comprising the following steps:

[0029] Step 1: Collect partial discharge information and create PRPD spectra

[0030] Collect the partial discharge signal on the valve side of the converter transformer, record the amplitude and phase of each partial discharge with T as the period, and draw the PRPD spectrum when the cumulative number of discharges exceeds N, and enter the second step. The collection method includes but is not limited to the pulse current method, ultra-high frequency method, ultrasonic method, etc., and the collection form includes but is not limited to continuous collection and triggered collection; T is an integer multiple of 20ms; N ≥ 100.

[0031] In this example: T = 20ms, N is 100, 500, 1000 respectively, the PRPD spectrum is as follows Figure 2 As shown in (a), (b) and (c).

[0032] Step 2: Delineate the PRPD spectrum

[0033] The method of describing the PRPD spectrum profile includes but is not limited to: dividing the PRPD spectrum into different phase regions at intervals of D, extracting the maximum discharge amplitude in each phase region, and sequentially connecting the maximum discharge amplitudes of each phase region to obtain a PRPD spectrum profile curve, wherein D is less than or equal to 5°.

[0034] In this example: D = 2°, the PRPD spectrum is cut vertically into 180 parts, and a curve of the maximum discharge amplitude versus phase is drawn to draw the PRPD spectrum profile curve. When N is 100, 500, and 1000, the PRPD spectrum profile curve is as follows: Figure 3 As shown in (a), (b) and (c).

[0035] Step 3: Frequency domain analysis of PRPD spectrum profile curve

[0036] In this example: Fast Fourier analysis is performed on each curve, and the results are as follows Figure 4 As shown in (a), (b) and (c).

[0037] Step 4: Identify main insulation and longitudinal insulation discharge

[0038] 1) If the amplitude of the 900Hz component accounts for the largest percentage of the fundamental frequency, it is longitudinal insulation discharge;

[0039] 2) If the 900Hz component is not the largest and the 100Hz component is greater than the 300Hz component, it is longitudinal insulation discharge;

[0040] 3) If the 900Hz component is not the largest and the 100Hz component is less than the 300Hz component, the main insulation discharge occurs.

[0041] In this example: Figure 4 From (a), (b), and (c), we can see that when N ≥ 100, the 100 Hz component in the fast Fourier analysis results is always smaller than the 300 Hz component, so partial discharge occurs in the main insulation.

Claims

1. A method for identifying main longitudinal insulation defects under complex electrical stress on the valve side of a converter transformer, characterized in that: The steps include: Step 1: Collect partial discharge information and make PRPD spectrum; Collect the partial discharge signal on the valve side of the converter transformer, record the amplitude and phase of each partial discharge with a period of T, and draw a PRPD spectrum when the cumulative number of discharges exceeds N; Step 2: Describe the PRPD spectrum profile and obtain the PRPD spectrum profile curve to show the change of the maximum discharge amplitude with the phase; Step 3: Perform frequency domain analysis on the PRPD spectrum profile curve; Step 4: Identify the main insulation and longitudinal insulation discharge; 1) If the amplitude of the 900Hz component accounts for the largest percentage of the fundamental frequency, it is longitudinal insulation discharge; 2) If the 900Hz component is not the largest and the 100Hz component is greater than the 300Hz component, it is longitudinal insulation discharge; 3) If the 900Hz component is not the largest and the 100Hz component is less than the 300Hz component, the main insulation discharge occurs.

2. The method for identifying main longitudinal insulation defects of converter transformer under complex electrical stress on valve side according to claim 1 is characterized in that: In the first step, the acquisition method for collecting the partial discharge signal on the valve side of the converter transformer adopts a pulse current method, a UHF method or an ultrasonic method, and the acquisition form adopts continuous acquisition or triggered acquisition.

3. The method for identifying main longitudinal insulation defects of converter transformer under complex electrical stress on valve side according to claim 1 is characterized in that: In the first step, T is an integer multiple of 20ms; N≥100.

4. The method for identifying main longitudinal insulation defects of converter transformer under complex electrical stress on valve side according to claim 1 is characterized in that: The method for characterizing the PRPD spectrum contour in the second step is: vertically cut the PRPD spectrum into different phase regions at intervals of D, extract the maximum discharge amplitude in each phase region, connect the maximum discharge amplitudes of each phase region in sequence, and obtain the PRPD spectrum contour curve.

5. The method for identifying main longitudinal insulation defects of converter transformer under complex electrical stress on valve side according to claim 4 is characterized in that: D is less than or equal to 5°.

6. The method for identifying main longitudinal insulation defects of converter transformer under complex electrical stress on valve side according to claim 1 is characterized in that: In the third step, the frequency domain analysis is performed using fast Fourier analysis.

Images