Power distribution network insulator early fault detection method and system considering interference event

Abstract

The invention provides a power distribution network insulator early fault detection method and system considering an interference event. The method comprises the steps of obtaining a zero-sequence current of a power distribution network insulator, and calculating a zero-sequence current mutation rate based on the zero-sequence current; if the zero-sequence current abrupt change rate is greater than a set threshold value, judging that the power distribution network has a fault or a disturbance event; obtaining original recording data of the insulator of the power distribution network, and filtering the original recording data based on a wavelet reconstruction algorithm; for the filtered signal, identifying a peak waveform based on a fault phase voltage extreme point; the non-linear distortion waveform is identified based on zero-sequence current sine fitting, and based on an identification result, when a set first criterion or a set second criterion is satisfied, it is judged that an early fault of the insulator may occur; and calculating the wavelet energy entropy of the signal, if the wavelet energy entropy is greater than a threshold value of the set entropy, determining that the power distribution network insulator has an early fault, otherwise, determining that the power distribution network insulator has an interference event.

Description

Technical Field

[0001] The present invention belongs to the technical field of early fault detection of insulators in a distribution network, and in particular to an early fault detection method and system for insulators in a distribution network taking interference events into consideration. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] The distribution network is an important part of the power system, and the reliability of its power supply is closely related to the safe production and life of electricity users. Improving the fault diagnosis and processing capabilities of the distribution network is an effective way to ensure the reliability of the distribution network power supply, shorten the power outage time and improve the user's power experience. With the increasing requirements for the reliability and safety of the distribution network power supply, the demand for the distribution network fault handling to change from "post-diagnosis" to "pre-warning" is becoming more and more urgent.

[0004] During the development of distribution network faults, the damage to the equipment generally changes from quantitative change to qualitative change. For example, under the action of operation or lightning overvoltage, the weak insulation points of the equipment are broken down and arcs are generated. Since the fault electrical parameters in the small current grounding system are small, they are not enough to trigger the protection action. In addition, the distribution network scenarios are complex and changeable, and the characteristics of non-fault operations and early faults are coupled, making it more difficult to accurately detect early faults.

[0005] For example, the early faults of insulators in distribution networks occur intermittently, and the electrical quantity characteristics are weak. The signals are submerged in a large number of disturbance events and are easily overlooked. The high temperature and thermal stress generated by intermittent arcs will cause irreversible damage to the surface of the porcelain body. The accumulated damage gradually transforms the early faults into permanent faults such as explosions and disconnections, causing serious impacts on the power grid and personnel. Therefore, it is necessary to analyze the early fault characteristics of insulators, discover and eliminate potential faults before the relay protection system is activated, and help reduce the existing fault handling workload.

[0006] The early faults of distribution network insulators are weak, and their characteristics vary according to the degree of insulation degradation and location. They are also easily confused by normal operation disturbances of the distribution network, such as capacitor switching and magnetizing inrush current. Therefore, the difficulty of early fault detection lies in sensitively detecting fault events while distinguishing them from interference events to ensure the reliability of fault detection. Summary of the invention

[0007] In order to overcome the above-mentioned deficiencies of the prior art, the present invention provides a distribution network insulator early fault detection method taking interference events into consideration, and performs early detection based on multi-dimensional features to improve the sensitivity and reliability of the detection method.

[0008] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:

[0009] In a first aspect, a method for early fault detection of insulators in a distribution network considering interference events is disclosed, comprising:

[0010] Obtaining the zero-sequence current of the insulator of the distribution network and calculating the zero-sequence current mutation rate based on the zero-sequence current;

[0011] If the zero-sequence current mutation rate is greater than the set threshold, it is judged that a fault or disturbance event has occurred in the distribution network;

[0012] Obtaining the original recorded data of the insulators of the distribution network and filtering the original recorded data based on the wavelet reconstruction algorithm;

[0013] For the filtered signal, the peak waveform is identified based on the extreme point of the fault phase voltage; the nonlinear distortion waveform is identified based on the zero-sequence current sine fitting, and based on the identification result, when the set first criterion or second criterion is met, it is determined that an early insulator fault may have occurred;

[0014] Then the wavelet energy entropy of the signal is calculated. If it is greater than the set entropy threshold, it is an early fault of the distribution network insulator, otherwise it is a interference event.

[0015] As a further technical solution, the filtered signal is evaluated by a curve fitting index, and the signal reconstruction result with the minimum index is used as the required fault feature extraction waveform.

[0016] As a further technical solution, the peak waveform is identified based on the extreme point of the fault phase voltage, specifically including: finding the extreme point of the fault phase voltage near the breakdown moment to identify the arcing spike, and then finding the next voltage extreme point. When the first criterion is met, it is considered to be an arcing spike.

[0017] As a further technical solution, when the first criterion is met, it is considered to be an arc extinction peak, and the criterion is specifically set as follows:

[0018]

[0019] Among them, D1 represents the extreme point at the breakdown moment, D h is the hth extreme point after breakdown; T D Corresponding to the extreme point, T is the power frequency period; A D is the amplitude of the extreme point; Used to indicate the increasing or decreasing trend of the arc.

[0020] As a further technical solution, the nonlinear distortion waveform is identified based on zero-sequence current sinusoidal fitting, specifically including:

[0021] The zero-sequence current data is divided into time windows of half the length of the power frequency cycle, and each segment of data takes the zero-sequence current point as the starting point;

[0022] For each point in the time window, a sine function is fitted to the zero-crossing point, and a total of N / 2 sine functions with different amplitudes are obtained;

[0023] The amplitudes of N / 2 sinusoidal functions in the time window tend to the maximum amplitude. When the amplitudes of the sinusoidal functions in the time window meet the second criterion, there is zero-sequence current nonlinear distortion.

[0024] As a further technical solution, the second criterion is specifically:

[0025] k1max(A mp (n)) <M(A mp (n)) <k2max(A mp (n))

[0026] Among them, A mp (n) is the set of amplitudes of each sine function in the time window, M(A mp (n)) is the median of the data, k1 and k2 are the lower and upper boundary coefficients.

[0027] As a further technical solution, interference event identification based on wavelet time-frequency energy entropy, the energy entropy index calculation process is:

[0028]

[0029] Among them, p j is the wavelet energy probability of each layer, D is the number of decomposition layers; C2 is the energy entropy index, and THR2 is the threshold obtained according to the experiment.

[0030] In a second aspect, a distribution network insulator early fault detection system considering interference events is disclosed, comprising:

[0031] The zero-sequence current mutation rate calculation module is configured to: obtain the zero-sequence current of the distribution network insulator and calculate the zero-sequence current mutation rate based on the zero-sequence current;

[0032] If the zero-sequence current mutation rate is greater than the set threshold, it is judged that a fault or disturbance event has occurred in the distribution network;

[0033] The recording data filtering module is configured to: obtain the original recording data of the distribution network insulator and filter the original recording data based on the wavelet reconstruction algorithm;

[0034] The waveform recognition module is configured to: for the filtered signal, identify the peak waveform based on the extreme point of the fault phase voltage; identify the nonlinear distortion waveform based on the zero-sequence current sine fitting, and based on the recognition result, when the set first criterion or second criterion is met, determine that an insulator early fault may have occurred;

[0035] The judgment module is configured to: then calculate the wavelet energy entropy of the signal, if it is greater than the set entropy threshold, it is an early fault of the distribution network insulator, otherwise it is an interference event.

[0036] One or more of the above technical solutions have the following beneficial effects:

[0037] Since the electrical quantity characteristics of early faults are weak, designing detection criteria based on fault waveform characteristics can improve detection sensitivity. In this process, it is necessary not only to filter the recorded data, but also to retain the slight change trend in the original data. Using wavelet transform to decompose the original recorded data into different frequency bands and designing the optimal index to retain the effective frequency band for reconstruction is a reliable method; considering the need to distinguish between faults and disturbance events, wavelet transform should not be limited to noise reduction processing. Using wavelet transform to perform further time-frequency analysis on the original recorded data and designing criteria based on the distribution law of time-frequency characteristics can effectively distinguish between faults and disturbance events.

[0038] The technical solution of the present invention analyzes the early fault waveform of the distribution network insulator based on the measured voltage and current data, and uses the waveform characteristics as the effective characteristics of early fault detection; considering the influence of short-term and long-term transient interference events, based on wavelet transform processing, the time-frequency characteristic distribution of faults and disturbances is analyzed, and accurate identification of early faults is achieved according to the characteristic differences.

[0039] Since there is noise in the measuring device and high-frequency arcing is prone to occur during the early stage of insulator faults, it is not conducive to the extraction of waveform features. The technical solution of the present invention takes the curve change trend as an indicator, uses wavelet reconstruction to denoise the fault phase voltage and zero-sequence current signals, and uses the denoised signals to extract waveform fault features to design the starting criterion.

[0040] The technical solution of the present invention combines the time-frequency distribution characteristics of early faults, extracts the energy of each frequency band after the zero-sequence current wavelet transform, and designs auxiliary criteria using the frequency band energy entropy through differentiated comparison with the time-frequency distribution of interference events; early fault detection is performed based on the multi-dimensional characteristics of the fault phase voltage, zero-sequence current waveform changes and zero-sequence current time-frequency energy distribution, thereby improving the sensitivity and reliability of the detection method.

[0041] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0043] Figure 1 This is a schematic diagram of recording early faults of insulators in a distribution network according to an embodiment of the present invention;

[0044] Figure 2 This is a schematic diagram of the waveforms of the fault phase voltage and zero-sequence current after filtering according to an embodiment of the present invention;

[0045] Figure 3 Schematic diagram of nonlinear distortion characteristics of zero-sequence current according to an embodiment of the present invention;

[0046] Figure 4 This is a schematic diagram of recording data of interference events in a power distribution system according to an embodiment of the present invention;

[0047] Figure 5 This is a schematic diagram of the time-frequency analysis results of fault and disturbance signals according to an embodiment of the present invention;

[0048] Figure 6 This is a schematic diagram of the time-frequency analysis results of fault and disturbance signals according to an embodiment of the present invention;

[0049] Figure 7 This is a flow chart of early fault detection in a distribution network according to an embodiment of the present invention. DETAILED DESCRIPTION

[0050] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0051] It should be noted that the terms used herein are for describing specific embodiments only and are not intended to be limiting of exemplary embodiments according to the present invention.

[0052] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0053] Analysis of early fault characteristics of distribution network insulators based on recorded data:

[0054] The development of early faults of insulators in distribution networks can be summarized into two common types on site. One is that dirt accumulates on the surface of the insulator after being exposed to wind and rain, thus forming a dirt surface flashover. The other is that under long-term mechanical loads and external environments with hot and cold changes, cracks appear on the insulators, and local arcs are generated after water vapor intrusion, which further evolves into a flashover.

[0055] Regardless of the type of early fault, the evolution process is similar. Due to the degradation of the insulation performance of the equipment, a high-resistance area appears in a certain part of the insulator, and the voltage at both ends is almost applied here. When the external applied voltage reaches the breakdown voltage, a local arc appears, and as the arc extends to the two poles, a flashover is formed.

[0056] For the fault current, the instantaneous current of breakdown changes suddenly, accompanied by high-frequency arcing. For the discharge along the surface of the insulator, the arc formed is relatively long, and the heat dissipation power is relatively large under the influence of external airflow; while the internal cracks of the insulator are narrow, and the creepage distance is much smaller than the external surface flashover, which mainly relies on the heat dissipation of the porcelain body, and can quickly develop from local to flashover. When the fault current drops to a certain value, the arc column diffuses too fast, and it is difficult to maintain the dynamic balance of the arc column diffusion, and arc extinction will occur; and the early faults of the insulator often occur near the voltage peak, and are mostly in the rising period, so the discharge duration is theoretically less than half a cycle.

[0057] It should be noted that in the early stage of the arc, the current change rate of the high-frequency current component is very large at the zero crossing point, and there is a possibility that the arc will not be extinguished even if the current is small; usually the arc current change rate inside the crack is relatively large, and no current interruption will occur during the high-frequency arcing period. After a short oscillation, the fault current is close to the power frequency current; while the external surface arc current change rate is even smaller, and current interruption often occurs when the high-frequency arcing passes through zero.

[0058] like Figure 1 The figure shows the waveform of the insulator early fault simulated on site; Figure 1 (a) and (c) are the zero-sequence current and fault phase voltage of the insulator external surface discharge respectively; Figure 1 (b) and (d) are the zero-sequence current and fault phase voltage of the internal crack discharge of the insulator. It can be observed that when the surface of the insulator is broken down, the initial arcing time is short and the overall current distortion is high; after the internal crack is broken down, the fault current is sinusoidal for a period of time after high-frequency arcing. At the same time, each time the breakdown and arcing are extinguished, the fault phase voltage has arcing spikes and arcing spikes.

[0059] Embodiment 1

[0060] This embodiment discloses a method for early fault detection of insulators in a distribution network considering interference events. Figure 7 As shown, including:

[0061] Step S1: using a fault recording device to collect fault phase voltage and zero-sequence current data;

[0062] Step S2: judging the collected data using the start criteria, if the start criteria are not met, returning to step S1, if the start criteria are met, proceeding to step S3;

[0063] Step S3: Determine whether the collected data meets the voltage criterion or the current criterion. If not, return to step S1. If yes, proceed to step S4.

[0064] Step S4: Calculate the wavelet energy entropy. If it is greater than a set threshold, it is judged as an early fault of the distribution network insulator. If it is less than a threshold, it is judged as a disturbance event.

[0065] In this implementation example, steps S2, S3, and S4 actually extract multiple fault features to achieve early fault detection and identification, and improve detection reliability through multi-level processing; step 3 designs voltage and current criteria based on the waveform characteristics of early faults, which are not conventional criteria, but are adjusted based on the waveform characteristics; step 4 designs criteria for the time-frequency energy distribution of early faults and system disturbances, and also makes adaptive adjustments. The three steps are similar to the sequential detection process of the flowchart. When all the criteria are met, it is judged as an early fault. The previous detection methods are mostly based on a single feature, and this embodiment extracts multiple fault features to jointly design detection criteria in order to improve detection reliability.

[0066] In step S2, the sub-technical solution of this embodiment uses the zero-sequence current mutation rate as the starting criterion. When the zero-sequence current mutation rate is greater than the threshold at a certain moment, it is considered that a fault or disturbance event has occurred in the distribution network, and the subsequent diagnosis process is started:

[0067] C1=di0 / dt>THR1 (1)

[0068] Among them, C1 is the mutation rate indicator, i0 is the zero-sequence current, and THR1 is the threshold value obtained according to the experiment.

[0069] The performance of insulator early faults varies according to the degree of insulation degradation. When the insulation is acceptable, the early fault is mainly intermittent discharge with a short duration; as the insulation damage deepens, the early fault is similar to a high-resistance grounding fault, and the fault current is a periodic waveform with distortion at the zero point. For example Figure 1 In (b), a destructive test is conducted on the insulator. The insulator is drilled vertically to the hardware. When the weather is clear, the insulator will not break down and discharge. However, when water invades the insulator, the early fault develops rapidly. After multiple breakdown discharges, a stable discharge channel is formed. Due to the continuous evaporation and invasion of water in this process, high-frequency arcing is exhibited at each zero-rest moment at the initial stage of the fault.

[0070] For intermittent arcs, the sub-technical scheme of this embodiment uses the current mutation value to detect the breakdown moment. When the fault phase voltage has both arcing peaks and arcing peaks within half a cycle, it is considered to be a valid fault feature. For early faults with fast and stable discharge, the zero-off period caused by the nonlinear distortion characteristics of the zero-sequence current is quantitatively extracted and used as a fault feature.

[0071] In step S3: before determining whether the collected data meets the voltage criterion or the current criterion, it is also necessary to extract the fault waveform features based on wavelet reconstruction:

[0072] The above method uses the current mutation rate during breakdown arcing as the starting criterion, but the high-frequency component and system noise will affect the extraction of subsequent fault phase voltage arcing spikes, arcing spikes and zero-sequence current zero-off period waveform features. The fault waveform feature extraction process of the present invention is as follows:

[0073] Step S3-1: Filter the original recorded data based on the wavelet reconstruction algorithm considering target optimization. It consists of two parts: wavelet decomposition and target optimization. The wavelet filtering algorithm is:

[0074]

[0075] Among them, S(t) is the original data, c j (t) is the jth component after wavelet transform, S F (t) is the filtered signal, which is obtained by optimizing the components to be filtered out. Since the subsequent fault characteristics are the quantization of the fault phase voltage spike waveform and the zero-sequence current zero-off period waveform, the filtered signal should focus on reflecting the change trend of the original data and establish the evaluation index:

[0076]

[0077] Among them, E r is the curve fitting index, N is the number of sampling points in one cycle; ΔS(t n ) and ΔS F (t n ) is the original signal and the filtered signal at each time t n Due to the presence of intermittent high-frequency noise in the early fault waveform, the degree of change is calculated by using the difference of the data mean in a short time window; the signal reconstruction result with the minimum index is used as the required fault feature extraction waveform. Figure 2 As shown, the fault phase voltage waveform and zero-sequence current waveform before and after filtering. After filtering, slight changes in the waveform can be seen.

[0078] Step S3-2: Identify the peak waveform based on the fault phase voltage extreme point.

[0079] After the fault detection program is started, the fault phase voltage extreme point is found near the moment of zero-sequence current mutation to identify the arcing spike, and then the next voltage extreme point is found. When the following criteria are met, it is considered to be an arcing spike:

[0080]

[0081] Among them, D1 represents the extreme point at the breakdown moment, D h is the hth extreme point after breakdown; T Dh 、T D1 is the breakdown moment and the moment of the hth extreme point after breakdown; AD1, ADh is the amplitude of the hth extreme point at the breakdown moment and after the breakdown; T is the power frequency period; It is used to indicate the increasing or decreasing trend of the arc. Formula (4) improves the accuracy of arc extinction peak recognition by combining the judgment method.

[0082] Step S3-3: Identify the nonlinear distortion waveform based on zero-sequence current sinusoidal fitting.

[0083] After the fault detection program is started, the zero-sequence current data is divided into time windows of half the length of the power frequency cycle, and each segment of data takes the zero-sequence current point as the starting point. For each point in the time window, a sine function is fitted to the zero-crossing point, and a total of N / 2 sine functions with different amplitudes are obtained, where N is the number of data sampling points in one cycle. According to the high-resistance fault resistance change curve, the equivalent fault resistance of the fault current is the smallest at the peak point and the largest near the zero-crossing point, and the fault resistance changes rapidly during the zero rest period. Therefore, the amplitudes corresponding to the N / 2 sine functions in the time window have an obvious distribution pattern, which can be attributed to the fact that the amplitudes of most sine functions tend to the maximum amplitude, such as Figure 3 shown.

[0084] When the amplitude of each sinusoidal function in the time window meets the following criteria, it is considered that there is zero-sequence current nonlinear distortion:

[0085] k1max(A mp (n)) <M(A mp (n)) <k2max(A mp (n)) (5)

[0086] Among them, A mp (n) is the set of amplitudes of each sine function in the time window, M(A mp (n)) is the median of the data, k1 and k2 are the lower and upper boundary coefficients, and this criterion is a targeted criterion proposed by the present invention based on the above nonlinear characteristics.

[0087] When either the voltage criterion (4) or the current criterion (5) is met, it is considered that an early insulator failure may have occurred.

[0088] In step S4, interference events are identified based on wavelet time-frequency energy entropy.

[0089] The early faults of insulators in the distribution network are self-recovering and hidden. At the signal level, they are similar to some system disturbance events. After the disturbance event occurs, it may cause the false triggering of the detection criteria. Figure 4 As shown, Figure 4 (a) in the figure is the on-site capacitor switching. Figure 4The waveforms of the magnetizing surge current recording in (b) belong to instantaneous transient interference and long-term transient interference respectively. The former has instantaneous discharge phenomenon, which is easily confused with intermittent arc, while the latter also has nonlinear distortion.

[0090] In order to ensure the safety of fault detection, the present invention uses wavelet transform to analyze the time-frequency characteristics of fault waveform and disturbance waveform, and designs a criterion to distinguish fault from interference from the perspective of frequency band energy distribution. The process is as follows:

[0091] Step S4-1: Signal time-frequency difference analysis.

[0092] The wavelet transform is used to perform time-frequency analysis on the zero-sequence current of fault and disturbance signals. The results are as follows: Figure 5 As shown, Figure 5 (a) is the analysis result of the zero-sequence current signal of the early fault. Figure 5 (b) is the analysis result of the capacitor switching zero-sequence current signal. Figure 5 (c) in the figure is the analysis result of the zero-sequence current signal of the excitation surge current.

[0093] It can be seen that at the moment of insulator early fault breakdown, the zero-sequence current shows a broadband distribution in the frequency domain, which is related to the arc burning-extinguishing in the early stage of the early fault; while the zero-sequence current of capacitor switching and excitation surge current is more concentrated in the frequency domain distribution.

[0094] Step S4-2: Design an energy entropy criterion to distinguish between faults and disturbances.

[0095] According to the above analysis, features can be extracted from the perspective of energy distribution to distinguish fault signals from disturbance signals, such as Figure 6 As shown in the figure, the energy distribution of early faults in the frequency domain is more chaotic.

[0096] The wavelet energy entropy index is designed in combination with information entropy. The more chaotic the information is, the greater the information entropy value is. In order to further amplify the energy entropy of early faults, a distribution coefficient is set for the energy of each frequency band. For system interference, since the energy distribution is relatively concentrated, the distribution coefficient has little effect on the calculation of energy entropy; for early faults, the uneven distribution coefficient will further increase the degree of chaos in energy distribution, which is conducive to the identification between faults and disturbances:

[0097]

[0098] Among them, p j is the wavelet energy probability of each layer; E j is the wavelet energy of each layer; i jis the j-th wavelet coefficient of the zero-sequence current; D is the number of wavelet coefficient layers, db4 is selected as the mother wavelet, and the zero-sequence current is decomposed into 7 layers, and 7 detail coefficients and 1 approximate coefficient can be obtained; λ(j) is the set distribution coefficient, which can be artificially set through field tests and is used to combine with the time-frequency characteristic distribution characteristics of the early fault of the present invention; C2 is the energy entropy index, and THR2 is the threshold value obtained according to the experiment.

[0099] It should be noted that although the wavelet energy and information entropy methods are commonly used in fault detection, the present invention performs time-frequency analysis on the field recorded waveforms of early faults and system disturbance events, and specifically designs wavelet energy entropy criteria; among them, the time-frequency energy difference analysis between early faults and other disturbances is not available in previous studies.

[0100] Overall, the flow chart of the early fault detection method for insulators in distribution networks is as follows: Figure 7 As shown, a multi-level fault diagnosis method is adopted to sensitively detect the occurrence of early faults while avoiding noise and interference events.

[0101] The present invention analyzes the on-site recording data of early-stage faults of insulators in the distribution network, extracts fault characteristics at the waveform angle and designs detection criteria; uses the current mutation rate as the starting criterion, and quantitatively extracts the arcing and arcing peak waveforms of the fault phase voltage and the nonlinear waveform characteristics of the zero-sequence current to identify possible early-stage faults.

[0102] The present invention analyzes the time-frequency characteristics of fault and disturbance signals through wavelet transformation, designs an energy entropy index from the perspective of energy distribution law, and is used to distinguish between fault and disturbance events; a multi-level fault diagnosis method is designed in combination with multi-dimensional fault characteristics to ensure the sensitivity and reliability of detecting weak fault signals when early faults occur.

[0103] Embodiment 2

[0104] The purpose of this embodiment is to provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the program.

[0105] Embodiment 3

[0106] The purpose of this embodiment is to provide a computer-readable storage medium.

[0107] A computer-readable storage medium stores a computer program, which executes the steps of the above method when executed by a processor.

[0108] Embodiment 4

[0109] The purpose of this embodiment is to provide a distribution network insulator early fault detection system considering interference events, including:

[0110] The zero-sequence current mutation rate calculation module is configured to: obtain the zero-sequence current of the distribution network insulator and calculate the zero-sequence current mutation rate based on the zero-sequence current;

[0111] If the zero-sequence current mutation rate is greater than the set threshold, it is judged that a fault or disturbance event has occurred in the distribution network;

[0112] The recording data filtering module is configured to: obtain the original recording data of the distribution network insulator and filter the original recording data based on the wavelet reconstruction algorithm;

[0113] The waveform recognition module is configured to: for the filtered signal, identify the peak waveform based on the extreme point of the fault phase voltage; identify the nonlinear distortion waveform based on the zero-sequence current sine fitting, and based on the recognition result, when the set first criterion or second criterion is met, determine that an insulator early fault may have occurred;

[0114] The judgment module is configured to: then calculate the wavelet energy entropy of the signal, if it is greater than the set entropy threshold, it is an early fault of the distribution network insulator, otherwise it is an interference event.

[0115] Embodiment 5

[0116] The purpose of this embodiment is to provide a computer program product containing instructions, which, when running on a computer, enables the computer to execute the methods and functions involved in any of the above embodiments.

[0117] The steps involved in the apparatus of the above embodiment correspond to the method embodiment 1, and the specific implementation method can refer to the relevant description part of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method in the present invention.

[0118] Those skilled in the art should understand that the modules or steps of the present invention described above can be implemented by a general-purpose computer device, or alternatively, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0119] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. A method for early fault detection of insulators in distribution networks considering interference events, characterized in that: include: Obtaining the zero-sequence current of the insulator of the distribution network and calculating the zero-sequence current mutation rate based on the zero-sequence current; If the zero-sequence current mutation rate is greater than the set threshold, it is judged that a fault or disturbance event has occurred in the distribution network; Obtaining the original recorded data of the insulators of the distribution network and filtering the original recorded data based on the wavelet reconstruction algorithm; For the filtered signal, the peak waveform is identified based on the extreme value point of the fault phase voltage; Based on the zero-sequence current sinusoidal fitting, the nonlinear distortion waveform is identified. When the identification result satisfies the first criterion or the second criterion, it is determined that an insulator early fault may have occurred. Then the wavelet energy entropy of the signal is calculated. If it is greater than the set entropy threshold, it is an early fault of the distribution network insulator, otherwise it is a interference event.

2. The method for early fault detection of insulators in a distribution network considering interference events as claimed in claim 1, characterized in that: The filtered signal is evaluated by the curve fitting index, and the signal reconstruction result with the minimum index is used as the required fault feature extraction waveform.

3. The method for early fault detection of insulators in a distribution network considering interference events as claimed in claim 1, characterized in that: The peak waveform is identified based on the extreme point of the fault phase voltage, specifically including: finding the extreme point of the fault phase voltage near the breakdown moment to identify the arcing spike, and then finding the next voltage extreme point, and when the first criterion is met, it is considered to be an arcing spike.

4. The method for detecting early faults of insulators in a distribution network considering interference events as claimed in claim 1, characterized in that: When the first criterion is met, it is considered to be an arc extinction peak, and the specific criterion is set as follows: Among them, D1 represents the extreme point at the breakdown moment, D h is the hth extreme point after breakdown; T D Corresponding to the extreme point, T is the power frequency period; A D is the amplitude of the extreme point; Used to indicate the increasing or decreasing trend of the arc.

5. The method for early fault detection of insulators in a distribution network considering interference events as claimed in claim 1, characterized in that: Identify nonlinear distortion waveform based on zero-sequence current sinusoidal fitting, including: The zero-sequence current data is divided into time windows of half the length of the power frequency cycle, and each segment of data takes the zero-sequence current point as the starting point; For each point in the time window, a sine function is fitted to the zero-crossing point, and a total of N / 2 sine functions with different amplitudes are obtained; The amplitudes of N / 2 sinusoidal functions in the time window tend to the maximum amplitude. When the amplitudes of the sinusoidal functions in the time window meet the second criterion, there is zero-sequence current nonlinear distortion.

6. The method for detecting early faults of insulators in a distribution network taking interference events into consideration according to claim 1, characterized in that: The second criterion is specifically: k1max(A mp (n))<M(A mp (n))<k2max(A mp (n)) Among them, A mp (n) is the set of amplitudes of each sine function in the time window, M(A mp (n)) median of the data, k1 and k2 are the lower and upper boundary coefficients; Preferably, based on the interference event identification of wavelet time-frequency energy entropy, the energy entropy index calculation process is: Among them, p j is the wavelet energy probability of each layer, D is the number of decomposition layers; C2 is the energy entropy index, and THR2 is the threshold obtained according to the experiment.

7. A distribution network insulator early fault detection system considering interference events, characterized in that: include: The zero-sequence current mutation rate calculation module is configured to: obtain the zero-sequence current of the distribution network insulator and calculate the zero-sequence current mutation rate based on the zero-sequence current; If the zero-sequence current mutation rate is greater than the set threshold, it is judged that a fault or disturbance event has occurred in the distribution network; The recording data filtering module is configured to: obtain the original recording data of the distribution network insulator and filter the original recording data based on the wavelet reconstruction algorithm; The waveform recognition module is configured to: for the filtered signal, identify the peak waveform based on the extreme value point of the fault phase voltage; Based on the zero-sequence current sinusoidal fitting, the nonlinear distortion waveform is identified. When the identification result satisfies the first criterion or the second criterion, it is determined that an insulator early fault may have occurred. The judgment module is configured to: then calculate the wavelet energy entropy of the signal, if it is greater than the set entropy threshold, it is an early fault of the distribution network insulator, otherwise it is an interference event.

8. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the method described in any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method described in any one of claims 1 to 6 are performed.

Images