CN-121978490-A - Extra-high voltage GIS breakdown point positioning method and system based on voltage wave head identification

CN121978490ACN 121978490 ACN121978490 ACN 121978490ACN-121978490-A

Abstract

The invention provides an extra-high voltage GIS breakdown point positioning method and system based on voltage wave head identification, which relate to the technical field of GIS breakdown point positioning and comprise the steps of obtaining transient voltage traveling waves generated during GIS breakdown and constructing double-end original voltage traveling wave time sequence data; the method comprises the steps of inputting traveling wave time sequence data into a wave head recognition model, carrying out self-adaptive noise reduction, multi-scale distortion contour recognition and feature fusion on the traveling wave time sequence through a front multi-scale feature extraction and cross-layer feature splicing channel of the wave head recognition model, outputting accurate fault wave head position indexes end to end, converting the moment that the traveling wave actually reaches the first end and the last end according to the fault wave head position indexes, calculating double-end absolute fault distances by using a traveling wave ranging mathematical model, obtaining the positions of fault breakdown points according to the double-end absolute fault distances, and realizing a high-precision GIS breakdown point positioning process. The GIS breakdown point positioning method and device can achieve GIS breakdown point positioning with low cost and high precision.

Inventors

MAO GUANGHUI
WANG KUNHAN
ZHANG XINGYU
PEI ZHEHAO
WEN TAO
SHI HAIPENG
WANG YANWEI
YU WEINAN
FAN XING
ZHANG PING
AN YIYAN
ZHOU YAN
DIAO FENGXIN

Assignees

国网内蒙古东部电力有限公司电力科学研究院
合肥工业大学

Dates

Publication Date: 20260505
Application Date: 20260409

Claims (10)

1. The extra-high voltage GIS breakdown point positioning method based on voltage wave head identification is characterized by comprising the following steps: Acquiring transient voltage traveling waves generated during GIS breakdown, and constructing double-end original voltage traveling wave time sequence data; Inputting traveling wave time sequence data into a wave head recognition model, performing self-adaptive noise reduction, multi-scale distortion contour recognition and feature fusion on the traveling wave time sequence through a front multi-scale feature extraction and cross-layer feature splicing channel of the wave head recognition model, and outputting accurate fault wave head position indexes end to end; Converting the time when the traveling wave actually reaches the first end and the last end according to the position index of the fault wave head, and calculating the absolute fault distance of the two ends by using a traveling wave ranging mathematical model; and obtaining the position of the fault breakdown point according to the absolute fault distance of the two ends, and realizing the positioning process of the GIS breakdown point with high precision.
2. The method for locating the breakdown point of the extra-high voltage GIS based on the voltage wave head identification according to claim 1, wherein the steps of obtaining the transient voltage traveling wave generated during the GIS breakdown and constructing the double-end original voltage traveling wave time series data comprise the following steps: the electro-optical crystal sensors arranged at the head end and the tail end of the pipeline are utilized to couple transient voltage traveling wave signals in real time; after receiving the unified time reference signal triggered by the high-precision synchronous time service module, the multichannel high-speed data acquisition card carries out synchronous analog-to-digital conversion on the double-end transient voltage traveling wave signal at a preset high-frequency sampling rate; Discrete data points in a time window set before and after the triggering moment are intercepted, and one-dimensional original voltage traveling wave time sequence data are formed.
3. The ultra-high voltage GIS breakdown point positioning method based on voltage wave head identification as claimed in claim 1, wherein the wave head identification model is a U-Net architecture, original voltage traveling wave time series data firstly enter a front-end multi-scale feature extractor to carry out front-end downsampling feature extraction, a common one-dimensional convolution layer in the front-end multi-scale feature extractor is replaced by a one-dimensional residual module, each residual module comprises two serially connected one-dimensional convolution layers, a cross-layer identity mapping path is added between the input and the output of the modules, the traveling wave time series data are input into an encoder, then feature mapping of the data is carried out in a first layer residual module, and an attention gating unit is connected in series in a cross-layer feature splicing channel which is originally directly connected, and the units simultaneously receive micro features from a front-end shallow layer and macro features from a network deep layer to carry out splicing fusion in the channel dimension.
4. The extra-high voltage GIS breakdown point positioning method based on voltage wave head recognition according to claim 3, wherein the feature map after splicing and fusion is subjected to mutation probability mapping calculation through a convolution layer at the tail end and a Softmax classifier, the feature map of multiple channels is compressed into a single-channel time sequence logic value by the convolution layer, the logic value sequence Z is input into the Softmax classifier, mapped into a mutation probability distribution sequence on the time sequence, finally index coordinates where the maximum probability value is located are directly searched on the generated one-dimensional probability distribution sequence, and an accurate wave head position index is output through Argmax functions.
5. The method for locating the breakdown point of the extra-high voltage GIS based on the voltage wave head identification according to claim 1, wherein the steps of converting the moment that the traveling wave actually reaches the first end and the last end according to the position index of the fault wave head and calculating the absolute fault distance of the two ends by using the traveling wave ranging mathematical model include the following steps: Converting the discrete wave head position index value into absolute physical time by combining the sampling frequency and the trigger zero point moment of the data acquisition card, wherein the specific conversion relation is that the head end absolute arrival time tm=T0+ Indexm/fs and the tail end absolute arrival time tn=T0+ Indexn/fs, wherein tm and tn are respectively converted head end and tail end absolute arrival times, T0 is the trigger zero point moment, fs is the sampling frequency, and Indexm and Indexn are respectively the head end and tail end wave head position index values; And substituting the absolute arrival time difference of the traveling waves at the first end and the last end into a classical double-end ranging physical formula, and calculating to obtain the absolute fault distance from the measuring end to the actual fault breakdown point.
6. The extra-high voltage GIS breakdown point positioning method based on voltage wave head identification according to claim 1, wherein the obtaining the position of the breakdown point according to the absolute fault distance of the two ends realizes the high-precision GIS breakdown point positioning process, and the method comprises the following steps: And (3) taking the absolute fault distance as a final fault breakdown point positioning result to carry out visual interface output, simultaneously comparing a large number of known real fault breakdown positions of the predicted calculation distance in an off-line training and shaping stage, calculating to obtain an absolute physical positioning error, taking the positioning error as a core feedback judgment index of a loss function, continuously calculating a network gradient through a back propagation algorithm, and iteratively adjusting all weights and bias parameters in a wave head recognition model.
7. Extra-high voltage GIS breakdown point positioning system based on voltage wave head discernment, its characterized in that includes: the signal acquisition module is used for acquiring transient voltage traveling waves generated during GIS breakdown and constructing double-end original voltage traveling wave time sequence data; The wave head recognition module is used for inputting the traveling wave time sequence data into the wave head recognition model, carrying out self-adaptive noise reduction, multi-scale distortion contour recognition and feature fusion on the traveling wave time sequence through a front-end multi-scale feature extraction and cross-layer feature splicing channel of the wave head recognition model, and outputting accurate fault wave head position indexes end to end; The distance conversion module is used for converting the moment that the traveling wave actually reaches the first end and the last end according to the position index of the fault wave head and calculating the absolute fault distance of the two ends by using the traveling wave ranging mathematical model; and the positioning feedback module is used for obtaining the position of the fault breakdown point according to the absolute fault distance of the two ends and realizing the positioning process of the GIS breakdown point with high precision.
8. A computer program product comprising a computer program, characterized in that the computer program, when executed by a processor, implements the extra-high voltage GIS breakdown point positioning method based on voltage wave head identification according to any one of claims 1-6.
9. A non-transitory computer readable storage medium for storing computer instructions which, when executed by a processor, implement the extra-high voltage GIS breakdown point positioning method based on voltage wave head identification of any one of claims 1-6.
10. An electronic device comprising a processor, a memory and a computer program, wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is operated, the processor executes the computer program stored in the memory, so that the electronic device executes the extra-high voltage GIS breakdown point positioning method based on voltage wave head identification according to any one of claims 1-6.

Description

Extra-high voltage GIS breakdown point positioning method and system based on voltage wave head identification Technical Field The disclosure relates to the technical field of GIS breakdown point positioning, in particular to an extra-high voltage GIS breakdown point positioning method and system based on voltage wave head identification. Background The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art. The gas-insulated metal-enclosed switchgear (GIS) has small occupied area, small influence from external environment and high reliability, and is widely applied to ultra-high voltage transmission systems. However, due to extremely high voltage level and complex internal electric field distribution of the extra-high voltage system, once insulation breakdown flashover faults occur in the GIS due to insulation aging, metal particles or foreign matters and the like, serious threat is caused to safe and stable operation of the power grid. Therefore, the breakdown point is rapidly and accurately positioned after the fault occurs, and the method is a key link for rapidly troubleshooting and recovering power supply. At present, ultrasonic positioning technology is often adopted in engineering practice to detect discharge and faults inside a GIS. The method mainly relies on capturing mechanical acoustic vibration signals generated at the moment of breakdown. However, in practical application, the ultrasonic positioning has obvious technical limitations that on one hand, when ultrasonic waves are transmitted in complex GIS internal physical structures (such as interfaces of different media, basin-type insulators and the like), serious attenuation and distortion of signals are easy to generate, so that a large error is generated in final physical space positioning, and on the other hand, the effective transmission distance of ultrasonic signals on a GIS metal shell is extremely limited. This results in the fact that in practical engineering applications, the ultrasonic probes must be densely arranged every very short length of GIS pipeline. This sensor configuration greatly increases hardware procurement costs, field installation effort, and later operational maintenance costs. In order to overcome the problems of low spatial resolution and high arrangement cost of ultrasonic positioning, a double-end traveling wave ranging method based on electromagnetic transient signals gradually becomes a more potential technical route. In theory, the two-end traveling wave positioning only needs to install one voltage sensor at the first end and the last end of a section of GIS bus respectively, and the accurate fault position is calculated by utilizing the time difference of the high-frequency voltage traveling wave generated in the breakdown moment reaching the sensors at the two ends and combining the traveling wave propagation speed. The method greatly reduces the number of sensors and the engineering cost. However, in a complex operating environment of an extra-high voltage GIS, the conventional double-end traveling wave ranging method faces the difficulty of accurate wave head extraction. When the initial voltage traveling wave generated at the moment of breakdown propagates in a GIS for a long distance, the initial voltage traveling wave can be influenced by factors such as skin effect, high-frequency loss of a gas medium, local impedance mutation and the like, and the high-frequency energy of the traveling wave can be severely attenuated and dispersed. This directly causes severe distortion of the otherwise steep transient traveling wave front, embodied by the rising and falling edges of the waveform becoming abnormally gentle. The traditional traveling wave positioning device generally adopts a fixed amplitude threshold value, a slope threshold value detection method or a pure mathematical analysis method such as wavelet transformation singular point calibration to find the arrival time of the traveling wave, but the traditional traveling wave positioning device still has the following limitations: (1) In the face of severe distortion, slow down and wave head signals accompanied by strong electromagnetic interference on site, the traditional methods can not accurately capture the real initial wave front, and misjudgment or missed judgment of the wave head starting point is very easy to occur. (2) Because the propagation speed of the electromagnetic traveling wave in the GIS is extremely fast (approaching to the light speed), even nanosecond-level wave head time calibration errors are amplified into a double-end ranging formula to be a few meters or even tens of meters of physical space positioning errors, and the ranging failure is directly caused. (3) The existing extra-high voltage GIS fault location technology is difficult to achieve balance between technical economy and ranging accuracy under complex working conditions. How t