CN-122017465-A - Power distribution network detection method, system and storage medium based on time domain signal correlation

CN122017465ACN 122017465 ACN122017465 ACN 122017465ACN-122017465-A

Abstract

The invention discloses a power distribution network detection method, a system and a storage medium based on time domain signal correlation, which belong to the technical field of power distribution network fault detection, can fully cover fault working conditions, break through the adaptation bottleneck of a static template library, completely cover nonlinear fault current characteristics of different grounding media, environmental humidity and fault initial conditions by constructing a normalized high-resistance grounding fault current mode group containing different ignition angles and random resistance parameters and matching with a special circuit model capable of simulating insulator leakage current characteristics, effectively solve the core defect that the static template library has limited scene adaptation range in the existing correlation detection technology, greatly improve the fault recognition adaptation capability under complex field working conditions, realize fault characteristic matching based on normalized cross correlation coefficient calculation, and effectively inhibit the influence of electromagnetic transient disturbance and background noise of the power distribution network under the environment of low signal-to-noise ratio by utilizing the inherent anti-noise characteristic of a correlation coefficient.

Inventors

QIN XIAOJUN
QIN TIANDAI
ZHANG KONG
FENG XIAOHUI
CHENG YUANHAO
GUO LIN
HAN BING
DU WENJUAN
CHEN LINGYU
ZHAO ZEXUAN
SONG ZHAOZHAO
XU GENLI
WEN YONG
SHI LIANG
WANG DANWEN
LONG JIE
YANG LILEI
ZHANG WEI

Assignees

国网河南省电力公司焦作供电公司
国网河南省电力公司修武县供电公司

Dates

Publication Date: 20260512
Application Date: 20260323

Claims (8)

1. The power distribution network detection method based on the time domain signal correlation is characterized by comprising the following steps of: S1, constructing a normalized high-resistance ground fault current mode family comprising different ignition angles and random resistances; S2, collecting a neutral point grounding wire current signal, filtering out a fundamental frequency component and a direct current component, and extracting a residual current signal; s3, calculating normalized cross correlation coefficients of the residual current signals and all normalized waveforms in the high-resistance ground fault current mode family; s4, setting a safety margin threshold, if the maximum correlation coefficient is in two continuous half periods If it is determined to be a high-resistance ground fault, if Then it is determined as a non-high resistance ground fault when Between 0.25 and 0.45, the half-cycle re-determination is delayed.
2. The method for detecting a power distribution network based on time domain signal correlation according to claim 1, wherein in S1, the range of ignition angles is: Wherein, the Is the ignition angle; in S1, the circuit model built by the high-resistance ground fault mode family is: a1, a series resistor and two anti-parallel branches, wherein each branch consists of a direct current source and a diode; A2, the leakage resistor is connected in parallel with the circuit model to simulate the leakage current characteristic of the insulator.
3. The method for detecting a power distribution network based on time domain signal correlation according to claim 1, wherein in S2, the neutral point ground line current signal is: the neutral point grounding wire current signal is collected as the sum of leakage currents of each phase of the line, the fundamental frequency component and the direct current component are filtered by a filter through the fact that the current of the neutral point of the grounding transformer contains fundamental harmonic waves and noise, and the residual current signal is obtained through extraction.
4. The method for detecting a power distribution network based on correlation of time domain signals according to claim 1, wherein in S3, the cross correlation coefficient is normalized The calculation formula of (2) is as follows: (1) In the formula (1) of the compound, For the number of sample points, For the pre-processed residual current signal, Normalizing the high-resistance ground fault current mode; In S3, the correlation coefficient is normalized: the robustness verification of the correlation coefficient is carried out according to the influence of noise on the correlation coefficient, which comprises that when the signal only contains fundamental frequency harmonic wave, the neutral point current is fundamental frequency component, which is uncorrelated with the high-resistance ground fault current mode family, and when the noise dominates the signal, the correlation coefficient The characteristics of (2) are: (2) In the formula (2) of the compound, Is the mean value of the two values, As a function of the variance of the values, Sampling points for each half cycle; when the signal contains high-resistance ground fault characteristics, the neutral point current signal is highly correlated with the high-resistance ground fault current mode family, the influence of noise is suppressed, and the variance of the correlation coefficient is: (3) in the formula (3) of the compound, For the signal-to-noise ratio, For the signal power to be high, Is the noise power.
5. The method for detecting a power distribution network based on time domain signal correlation according to claim 1, wherein in S4, the safety margin threshold is set as follows: Through multiple experimental tests on dry soil and grasslands To determine the lower threshold of high-resistance ground fault, the low threshold is determined by transformer inrush current and load imbalance non-high-resistance ground fault test To determine the upper threshold of the non-high-resistance ground fault, when Between 0.25 and 0.45, and re-judging after the subsequent half period; in S4, the deferred determination mechanism specifically includes: When (when) Between 0.25 and 0.45, the correlation coefficient of the two subsequent half cycles should be continuously monitored, if any And judging the high-resistance ground fault, otherwise judging the non-high-resistance ground fault.
6. The power distribution network detection system based on the time domain signal correlation is characterized by comprising a current transformer, a filter module, an embedded processor and an alarm module: The current transformer is arranged at the neutral point zigzag grounding connection part and is used for collecting neutral point current signals; the filter module is used for filtering fundamental frequency harmonic waves and direct current components; The embedded processor performs correlation coefficient calculation and threshold judgment, wherein the pre-stored high-resistance ground fault mode family database comprises high-resistance ground fault current waveform parameters corresponding to different surface humidities and contact materials; and the alarm module is used for triggering an alarm signal when the high-resistance ground fault is detected.
7. The system for power distribution network detection based on time domain signal correlation as recited in claim 6, further comprising a computer device comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, and wherein the processor implements the method for power distribution network detection based on time domain signal correlation when executing the computer program.
8. The power distribution network detection method based on the time domain signal correlation is characterized in that a computer program is stored on the storage medium, and when the computer program is executed by a processor, the method is used for realizing any one of the power distribution network detection methods based on the time domain signal correlation.

Description

Power distribution network detection method, system and storage medium based on time domain signal correlation Technical Field The invention relates to the technical field of power distribution network fault detection, in particular to a power distribution network detection method, a power distribution network detection system and a storage medium based on time domain signal correlation. Background The distribution network is used as an end transmission network directly facing end users in a power system, and the power supply reliability and the operation safety of the distribution network are directly related to the stable and orderly development of social production and life. The high-resistance ground fault is a typical fault type in the running process of the power distribution network, is mostly caused by the fact that a power distribution line wire contacts with high-impedance mediums such as dry soil, tree branches and concrete buildings, the fault current amplitude of the fault is extremely small, and the current waveform is characterized by strong nonlinearity and strong randomness, so that the conventional protection means are difficult to effectively identify. According to the statistical data of industries, about 30% of power failure accidents in a power distribution network are all caused by the fact that high-resistance ground faults cannot be timely and accurately detected, serious personal electric shock potential safety hazards can be caused by continuous existence of faults, and serious safety accidents such as electric fires and burning of power distribution equipment are extremely easy to occur, so that immeasurable economic losses and social effects are caused. At present, the main stream detection technology of the high-resistance ground fault of the power distribution network is mainly divided into three categories of a traditional relay protection method, a frequency domain harmonic analysis method and a time-frequency transformation analysis method. The method is easy to be interfered by working conditions such as nonlinear load switching and capacitor bank start-stop in a power distribution network, the fault false detection rate is high, and the time-frequency transformation analysis method represented by wavelet transformation can capture transient signal characteristics when faults occur, but the method has high calculation complexity, is difficult to meet the real-time requirement of power distribution network fault detection, has high sensitivity to the setting of a mother wavelet function selection and a signal decomposition layer number, and has serious inapplicability in a complex field operation environment. In recent years, a high-resistance ground fault detection technology based on signal correlation analysis gradually becomes an industry research hotspot, and the core principle is that a fault waveform template library is pre-constructed, actual measurement signals acquired on site are subjected to waveform matching with the template library, and fault identification is completed based on signal correlation. The existing detection technology based on correlation analysis still has obvious technical bottlenecks that firstly, a fault mode library is a static template library, dynamic fault characteristics under working conditions such as different fault ignition angles, line contact materials, on-site environment humidity and the like cannot be covered on the whole surface, the adaptation range of a fault scene is limited, secondly, the anti-noise and anti-interference capabilities of an algorithm are insufficient, the background noise can cause great fluctuation of a correlation coefficient under common working conditions such as unbalanced load and electromagnetic transient disturbance of a power distribution network, the reliability of fault detection is directly reduced, thirdly, a fault judging mechanism is single, a fixed threshold is commonly adopted to finish judgment, and non-fault working conditions such as high-resistance ground faults, transformer excitation surge, capacitor switching and the like with similar waveform characteristics cannot be effectively distinguished, so that the accuracy and the anti-misoperation capability of fault detection are difficult to be considered. The method, the system and the storage medium for detecting the power distribution network based on the time domain signal correlation are provided for solving the core pain points of low detection precision, weak anti-interference capability, poor scene suitability, insufficient instantaneity and the like in the existing power distribution network high-resistance ground fault detection technology. Disclosure of Invention The invention aims to provide a power distribution network detection method, a system and a storage medium based on time domain signal correlation, so as to solve the problems in the background art. In order to achieve the above purpose, the invention provides a power