Ultrasonic rapid location method for partial discharge in GIS based on pattern recognition and medium

Abstract

The present invention relates to a method and medium for rapid ultrasonic localization of partial discharge in a GIS (GIS) based on pattern recognition. The method comprises the following steps: constructing a GIS simulation model that matches the GIS device to be identified, obtaining an acoustic simulation fingerprint library based on the GIS simulation model; extracting the on-site partial discharge acoustic fingerprint of the GIS device to be identified, and determining the location of the partial discharge source based on pattern recognition results between the on-site partial discharge acoustic fingerprint and the acoustic simulation fingerprint library. Compared with existing technologies, the present invention offers advantages such as rapidity and accuracy.

Description

Technical Field

[0001] The present invention belongs to the field of electric power technology and relates to a method for locating a local discharge source, in particular to a method and medium for quickly locating a GIS local discharge using ultrasonic waves based on pattern recognition. Background Art

[0002] Rapidly and accurately locating partial discharge (PD) sources within enclosed switchgear (GIS) equipment allows for rapid assessment of internal insulation conditions, effectively reducing maintenance time and power outage costs. Ultrasonic detection of partial discharge (PD) has gained widespread application due to its immunity to electromagnetic interference, non-destructive nature, and non-invasive nature. Current approaches to locating PD sources within GIS equipment suffer from low accuracy and long location times. Existing PD location methods can be categorized into two types: those based on time difference of arrival (TDOA) and those based on spatial spectrum directionality. In terms of TDOA method, the traditional method is to solve the spatial spherical or hyperbolic equations through the least squares method and Newton iteration method. The current main method for positioning is intelligent algorithm, such as applying genetic algorithm to positioning, using particle swarm algorithm to solve the partial discharge positioning optimization problem, and using path search positioning method based on particle swarm optimization. However, these methods all need to assume that the sound speed of ultrasound is a constant, which does not conform to the propagation characteristics of ultrasound in GIS equipment, and therefore will cause certain positioning errors; the spatial spectrum-based orientation methods mainly include high-resolution spectrum estimation method and multiple signal classification (MUSIC) algorithm. Its orientation principle is to use the orthogonality of signal subspace and noise subspace for orientation, but it has high requirements on signal-to-noise ratio. When the signal-to-noise ratio is lower than 5dB, the noise will destroy its orthogonality, resulting in low orientation accuracy. Summary of the Invention

[0003] The purpose of the present invention is to overcome the problems of low accuracy and long positioning time in the above-mentioned prior art of GIS equipment internal partial discharge source positioning, and to provide a fast and accurate GIS partial discharge ultrasonic rapid positioning method and medium based on pattern recognition.

[0004] The purpose of the present invention can be achieved by the following technical solutions:

[0005] A GIS partial discharge ultrasonic rapid positioning method based on pattern recognition includes the following steps:

[0006] Build a GIS simulation model that matches the GIS device to be identified, and obtain an acoustic simulation fingerprint library based on the GIS simulation model;

[0007] The on-site partial discharge acoustic fingerprint of the GIS equipment to be identified is extracted, and the location of the partial discharge source is determined based on the pattern recognition results of the on-site partial discharge acoustic fingerprint and the acoustic simulation fingerprint library.

[0008] Furthermore, the size and model of the GIS simulation model are the same as those of the GIS device to be identified.

[0009] Furthermore, the acoustic simulation fingerprint library is obtained by the following steps:

[0010] Arranging a plurality of detection points on the cavity surface of the GIS simulation model;

[0011] Conducting partial discharge ultrasonic simulation tests at different locations in the cavity, obtaining sound pressure time-domain response maps containing partial discharge source coordinate information from the multiple detection points, wherein a corresponding sound pressure time-domain response map is obtained for each detection point during each partial discharge ultrasonic simulation test;

[0012] Feature extraction is performed on the sound pressure time domain response spectrum, and the sound pressure time domain response spectrum is converted into a feature vector. Each local discharge source coordinate information corresponds to a feature vector, and multiple feature vectors are constructed to obtain the acoustic simulation fingerprint library.

[0013] Furthermore, the characteristic parameters of the feature extraction include signal mutation point, maximum peak point, pulse rise time, pulse duration and amplitude.

[0014] Furthermore, after the feature extraction is performed, the extracted first vector is normalized to obtain the feature vector.

[0015] Furthermore, one acoustic simulation fingerprint library is constructed corresponding to one GIS device to be identified.

[0016] Furthermore, the extraction of the on-site partial discharge acoustic fingerprint of the GIS device to be identified specifically includes:

[0017] Acquiring a time-domain sound pressure signal collected by an ultrasonic sensor, wherein the ultrasonic sensor is arranged on the GIS device to be identified according to the detection point;

[0018] Preprocessing the time-domain sound pressure signal;

[0019] The pre-processed time-domain sound pressure signal is subjected to feature extraction to obtain the on-site real-time vector, which is the on-site partial discharge acoustic fingerprint.

[0020] Furthermore, the number of the ultrasonic sensors is less than or equal to the number of detection points.

[0021] Furthermore, the preprocessing includes denoising and filtering.

[0022] The present invention also provides a computer-readable storage medium, characterized in that it includes one or more programs for execution by one or more processors of an electronic device, and the one or more programs include instructions for executing the above-mentioned GIS partial discharge ultrasonic rapid positioning method.

[0023] Compared with the prior art, the present invention has the following beneficial effects:

[0024] 1. The present invention establishes a partial discharge acoustic simulation fingerprint library through finite element simulation. Through pattern recognition of the fingerprint library and the on-site partial discharge acoustic fingerprint measured on site, the partial discharge source position can be quickly located with high reliability.

[0025] 2. The present invention uses pattern recognition to achieve positioning, and the recognition accuracy is high. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a positioning flow chart of the present invention;

[0027] Figure 2 An example of the arrangement of detection points on the cavity surface and partial discharge sources inside the constructed GIS simulation model. DETAILED DESCRIPTION

[0028] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0029] This embodiment provides a method for rapid ultrasonic localization of GIS partial discharge based on pattern recognition, comprising the following steps: constructing a GIS simulation model that matches the GIS device to be identified, and obtaining an acoustic simulation fingerprint library based on the GIS simulation model; extracting the on-site partial discharge acoustic fingerprint of the GIS device to be identified, and determining the location of the partial discharge source based on the pattern recognition results of the on-site partial discharge acoustic fingerprint and the acoustic simulation fingerprint library.

[0030] like Figure 1 As shown, the above method mainly includes three parts: establishing an acoustic simulation fingerprint library, extracting on-site partial discharge acoustic fingerprints, and pattern recognition.

[0031] 1) Establish an acoustic simulation fingerprint library

[0032] The acoustic simulation fingerprint library is obtained by the following steps: arranging multiple detection points on the cavity surface of the GIS simulation model; performing partial discharge ultrasonic simulation tests at different positions in the cavity, obtaining sound pressure time domain response maps containing partial discharge source coordinate information from the multiple detection points, and obtaining a corresponding sound pressure time domain response map for each detection point in each partial discharge ultrasonic simulation test; performing feature extraction on the sound pressure time domain response map, converting the sound pressure time domain response map into a feature vector, wherein each partial discharge source coordinate information corresponds to a feature vector, and constructing the acoustic simulation fingerprint library using multiple feature vectors.

[0033] In this embodiment, Figure 2 As shown in the figure, a GIS simulation model was constructed based on the same size and dimensions as the on-site GIS equipment. Furthermore, s detection points were placed on the surface of the GIS simulation model cavity. The PD acoustic simulation fingerprint library can be constructed based on different equipment models, making it more suitable for GIS equipment of different sizes and models, and has a wide range of applications.

[0034] Based on the above simulation model, ultrasonic simulations of partial discharges were conducted at various locations within the cavity. Time-domain sound pressure response maps were obtained from s detection points distributed across the GIS cavity surface. N partial discharge simulations were conducted within the GIS simulation model, and each test yielded s time-domain sound pressure response maps. This resulted in a total of s × N time-domain sound pressure response maps containing information about the coordinates of the partial discharge sources.

[0035] The time-domain response spectrum of the sound pressure collected at the detection point is feature extracted and represented as a low-dimensional vector. This can greatly reduce the storage space occupied by the parameters, speed up the calculation rate, and facilitate the rapid positioning of the partial discharge point.

[0036] The characteristic parameters commonly used in the analysis and processing of ultrasonic time domain signals include signal mutation point, maximum peak point, pulse rise time, pulse duration and amplitude. Select the above p characteristic parameters and express the sound pressure time domain response spectrum collected at s detection points as a vector K of length s×p i =[k i,1 k i,2 …k i,j …k i,s ], corresponding to a local discharge point in the internal space of the GIS cavity. i,j =[m1,m2,…,m p ] is the characteristic parameter corresponding to the time-domain response spectrum of the PD acoustic pressure collected at the jth detection point (1≤j≤s). To highlight the strength differences and distribution patterns of the same PD source signal between sensors, the vectors extracted from the PD source features at the same location are normalized to [-1, 1].

[0037] All the above-mentioned sound pressure time-domain response spectra are feature extracted and normalized, thereby constructing a partial discharge acoustic simulation fingerprint library K containing N partial discharge source location information.

[0038]

[0039] Each row in this database corresponds to a partial discharge source in the GIS simulation model. This PD acoustic simulation fingerprint library can be constructed based on different device models, making it more adaptable to GIS equipment of varying sizes and models, thus offering broad application value. The placement of on-site ultrasonic sensors is based on the location of detection points on the cavity surface in the simulation model, and the number of ultrasonic sensors can be smaller than the number of detection points in the simulation model.

[0040] 2) Extraction of on-site partial discharge acoustic fingerprints

[0041] For the GIS equipment on site, the ultrasonic sensors are arranged according to the number and positions of the detection points set in the simulation model, and the number of ultrasonic sensors is recorded as q (q≤s).

[0042] Since the partial discharge ultrasonic signals measured on site often contain various noise interferences, such as white noise interference and periodic narrowband interference, in order to prevent interference, the first step is to pre-process the time-domain sound pressure signals collected by all ultrasonic sensors by denoising and filtering.

[0043] For the preprocessed time-domain sound pressure signal, the same feature parameters as those used to establish the acoustic simulation fingerprint library are selected to extract features. The number of selected feature parameters is denoted as p, and the time-domain sound pressure signal collected by q ultrasonic sensors is represented as a p×q column feature vector X=[x1,x2,…x j …,x q ], where x j =[m1,m2,…,m p ] represents the characteristic parameter corresponding to the PD acoustic pressure signal collected by the jth sensor. To highlight the strength differences and distribution patterns between sensors, the characteristic vector X is normalized, thus completing the extraction of the on-site PD acoustic fingerprint.

[0044] 3) Implementation of GIS partial discharge positioning

[0045] Perform pattern matching on-site partial discharge acoustic fingerprint X with the data in the acoustic simulation fingerprint library K, and identify the fingerprint vector K in the i-th row of the fingerprint library. i =[k i,1 k i,2 … k i,36], the simulated position of the PD source corresponding to the fingerprint is the actual position of the PD point in GIS.

[0046] If the above method is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program code.

[0047] The above describes in detail the preferred embodiments of the present invention. It should be understood that those skilled in the art can make numerous modifications and variations based on the concepts of the present invention without inventive effort. Therefore, any technical solutions that can be derived by those skilled in the art through logical analysis, reasoning, or limited experimentation based on the concepts of the present invention and the prior art should be within the scope of protection defined by the claims.

Claims

1. A GIS partial discharge ultrasonic rapid positioning method based on pattern recognition, characterized in that: The following steps are involved: Build a GIS simulation model that matches the GIS device to be identified, and obtain an acoustic simulation fingerprint library based on the GIS simulation model; Extracting an on-site partial discharge acoustic fingerprint of the GIS equipment to be identified, and determining the location of the partial discharge source based on a pattern recognition result of the on-site partial discharge acoustic fingerprint and the acoustic simulation fingerprint library; The acoustic simulation fingerprint library is obtained by the following steps: Arranging a plurality of detection points on the cavity surface of the GIS simulation model; Conducting partial discharge ultrasonic simulation tests at different locations in the cavity, obtaining sound pressure time-domain response maps containing partial discharge source coordinate information from the multiple detection points, wherein a corresponding sound pressure time-domain response map is obtained for each detection point during each partial discharge ultrasonic simulation test; Performing feature extraction on the sound pressure time domain response spectrum, converting the sound pressure time domain response spectrum into a feature vector, wherein each local discharge source coordinate information corresponds to a feature vector, and a plurality of the feature vectors are used to construct the acoustic simulation fingerprint library; The characteristic parameters of the feature extraction include signal mutation point, maximum peak point, pulse rise time, pulse duration and amplitude.

2. The GIS partial discharge ultrasonic rapid location method based on pattern recognition according to claim 1 is characterized in that: The size and model of the GIS simulation model are the same as those of the GIS device to be identified.

3. The GIS partial discharge ultrasonic rapid location method based on pattern recognition according to claim 1 is characterized in that: After the feature extraction is performed, the extracted first vector is normalized to obtain the feature vector.

4. The GIS partial discharge ultrasonic rapid location method based on pattern recognition according to claim 1 is characterized in that: An acoustic simulation fingerprint library is constructed corresponding to each GIS device to be identified.

5. The GIS partial discharge ultrasonic rapid location method based on pattern recognition according to claim 3 is characterized in that: The extraction of the on-site partial discharge acoustic fingerprint of the GIS device to be identified specifically includes: Acquiring a time-domain sound pressure signal collected by an ultrasonic sensor, wherein the ultrasonic sensor is arranged on the GIS device to be identified according to the detection point; Preprocessing the time-domain sound pressure signal; The pre-processed time-domain sound pressure signal is subjected to feature extraction to obtain the on-site real-time vector, which is the on-site partial discharge acoustic fingerprint.

6. The GIS partial discharge ultrasonic rapid location method based on pattern recognition according to claim 5 is characterized in that: The number of the ultrasonic sensors is less than or equal to the number of detection points.

7. The GIS partial discharge ultrasonic rapid location method based on pattern recognition according to claim 5 is characterized in that: The preprocessing includes denoising and filtering.

8. A computer-readable storage medium, characterized in that It comprises one or more programs for execution by one or more processors of an electronic device, wherein the one or more programs include instructions for executing the GIS partial discharge ultrasonic rapid location method as claimed in any one of claims 1-7.

Images