CN-122017484-A - Transformer substation partial discharge positioning and imaging method based on acoustic phased array

CN122017484ACN 122017484 ACN122017484 ACN 122017484ACN-122017484-A

Abstract

The invention relates to the field of insulation monitoring and fault positioning of power equipment, in particular to a transformer substation partial discharge positioning and imaging method based on an acoustic phased array. The method comprises the steps of deploying an acoustic phased array sensor array to collect local discharge multichannel original acoustic signals, obtaining array frequency domain signal data through noise suppression and signal enhancement pretreatment, calculating a spatial spectrum through a self-adaptive beam forming algorithm, separating multiple discharge source candidate space positions through peak detection and cluster analysis, accurately positioning by using a time difference positioning principle to obtain three-dimensional space coordinates, and fusing energy intensity information to generate a local discharge three-dimensional space distribution image. The method effectively solves the core problems of limited dimension of partial discharge positioning, lack of visual imaging function, poor separation effect of multiple discharge sources and insufficient anti-interference capability in the prior art.

Inventors

MA ZIHAN
DONG LILIANG
LIU SHUAI
DONG ZHIJIANG
LI ZHIQIANG
DUAN MIN
YUAN ZHIYONG
ZHANG YALI

Assignees

国网河北省电力有限公司临漳县供电分公司
国网河北省电力有限公司邯郸供电分公司
国家电网有限公司

Dates

Publication Date: 20260512
Application Date: 20260126

Claims (10)

1. A transformer substation partial discharge positioning and imaging method based on an acoustic phased array is characterized by comprising the steps of deploying an acoustic phased array sensor array in a transformer substation monitoring area, collecting multichannel original acoustic signals generated by partial discharge to obtain array time domain signal data, preprocessing the array time domain signal data, including noise suppression and signal enhancement, to obtain preprocessed array frequency domain signal data, calculating a spatial spectrum through an adaptive beam forming algorithm based on the preprocessed array frequency domain signal data to obtain initial spatial spectrum distribution containing energy peaks, carrying out peak detection and cluster analysis on the energy peaks in the initial spatial spectrum distribution, separating out a plurality of candidate spatial positions corresponding to different discharge sources, carrying out accurate positioning calculation on each candidate spatial position by utilizing a time difference positioning principle and the array time domain signal data to obtain three-dimensional space coordinates of each discharge source, and fusing the three-dimensional space coordinates of the discharge sources and energy intensity information of the discharge sources in the initial spatial spectrum distribution to generate a partial discharge three-dimensional spatial distribution image of the transformer substation monitoring area.
2. The substation partial discharge positioning and imaging method based on the acoustic phased array according to claim 1, wherein the step of preprocessing the array time domain signal data to obtain preprocessed array frequency domain signal data specifically comprises the following steps: The method comprises the steps of carrying out wavelet threshold noise reduction processing on array time domain signal data of each channel to obtain array time domain signal data after noise reduction, carrying out self-adaptive blind source separation on the array time domain signal data after noise reduction to separate background noise components and potential discharge acoustic signal components to obtain preliminary separated acoustic signals, carrying out band-pass filtering on the preliminary separated acoustic signals, reserving a frequency band matched with partial discharge acoustic emission characteristics to obtain filtered acoustic signals, carrying out short-time Fourier transform on the filtered acoustic signals to convert the signals from time domain to obtain time-frequency domain signal representation, further enhancing the time-frequency domain signal representation by using a wiener filtering algorithm in the time-frequency domain to inhibit residual non-stationary noise to obtain enhanced time-frequency domain signals, carrying out frequency domain integration on the enhanced time-frequency domain signals to extract energy distribution in a specific frequency band, and finally obtaining preprocessed array frequency domain signal data.
3. The substation partial discharge positioning and imaging method based on the acoustic phased array according to claim 1, wherein the step of calculating a spatial spectrum through an adaptive beam forming algorithm based on the preprocessed array frequency domain signal data to obtain an initial spatial spectrum distribution specifically comprises the following steps: The method comprises the steps of establishing an accurate geometric position model of an acoustic phased array sensor array, calculating theoretical acoustic wave propagation delay from each point to be scanned to each array element in a monitoring area according to the geometric position model to form an array flow pattern matrix, carrying out inversion operation on a covariance matrix of preprocessed array frequency domain signal data by adopting a minimum variance undistorted response beam forming algorithm, calculating the spatial power spectral density of each scanning point by combining the array flow pattern matrix, carrying out regularization treatment on the covariance matrix by introducing a diagonal loading technology to improve the numerical stability of the algorithm in a coherent signal environment, and carrying out logarithmic compression and normalization treatment on the calculated original spatial power spectral density to form visualized initial spatial spectrum distribution.
4. The method for positioning and imaging partial discharge of a transformer substation based on acoustic phased array according to claim 1, wherein the step of performing peak detection and cluster analysis on energy peaks in an initial spatial spectrum distribution to separate a plurality of candidate spatial positions corresponding to different discharge sources specifically comprises: Setting a dynamic energy threshold in initial spatial spectrum distribution, extracting all local maximum points higher than the dynamic energy threshold to obtain an initial peak point set, calculating three-dimensional space coordinates and energy values of each point in the initial peak point set to form a three-dimensional space energy point cloud, analyzing the three-dimensional space energy point cloud by using a density-based spatial clustering algorithm, merging peak points with adjacent space positions and similar energy values into the same cluster to obtain a plurality of candidate peak clusters, calculating an energy weighted centroid of each candidate peak cluster, taking the three-dimensional space coordinates of the centroid as the candidate space position of one candidate discharge source represented by the cluster, removing the invalid candidate peak clusters with energy sum lower than the set threshold, and outputting the plurality of candidate space positions finally separated and corresponding energy information.
5. The method for positioning and imaging partial discharge of a transformer substation based on an acoustic phased array according to claim 1, wherein the step of precisely positioning and calculating each candidate space position by using a time difference positioning principle and array time domain signal data to obtain three-dimensional space coordinates of each discharge source specifically comprises the following steps: According to each candidate space position, selecting time domain signals of a plurality of sensor channels with strongest energy from noise-reduced array time domain signal data as reference signals, calculating cross-correlation functions between the time domain signals of the rest sensor channels and the reference signals, precisely estimating time difference of sound waves transmitted to different array elements by searching global maximum points of the cross-correlation functions to obtain a high-precision time difference estimated value, establishing a time difference positioning equation set taking the candidate space position as an initial solution according to a precise geometric position model of an acoustic phased array sensor array and a propagation speed model of the sound waves in a complex environment of a transformer substation, solving the time difference positioning equation set by adopting an iterative least square algorithm until residual errors between the time difference estimated value and theoretical time difference values calculated according to current coordinates are minimized, and finally determining the optimized three-dimensional coordinates as a precise positioning result of the power supply.
6. The method for positioning and imaging partial discharge of a transformer substation based on an acoustic phased array according to claim 1, wherein the step of generating a partial discharge three-dimensional spatial distribution image of a monitoring area of the transformer substation by fusing three-dimensional spatial coordinates of a discharge source and energy intensity information thereof in initial spatial spectrum distribution specifically comprises: The method comprises the steps of constructing a three-dimensional voxel space grid covering a transformer substation monitoring area, mapping three-dimensional space coordinates of each discharge source obtained through accurate positioning into the three-dimensional voxel space grid, finding a voxel unit where the discharge source is located, assigning corresponding gray values or color values to the voxel unit according to energy intensity information corresponding to the discharge source in initial space spectrum distribution, interpolating and filling the voxel unit which is not directly occupied by the discharge source by adopting an inverse distance weighted interpolation algorithm based on distance, according to values of adjacent discharge source voxels, forming a continuous energy intensity space distribution field, and performing volume drawing on energy intensity data in the three-dimensional voxel space grid by using a three-dimensional image rendering technology, so as to generate a transformer substation partial discharge three-dimensional space distribution image containing space position and discharge intensity information.
7. The substation partial discharge positioning and imaging method based on the acoustic phased array according to claim 1, further comprising a discharge type auxiliary identification step: Extracting array time domain signal data fragments corresponding to a precisely positioned discharge source to obtain target discharge signal fragments, extracting features of the target discharge signal fragments, calculating time domain waveform features, frequency domain spectral line features and time frequency joint features of the target discharge signal fragments to form discharge acoustic pattern feature vectors, inputting the discharge acoustic pattern feature vectors into a pre-trained discharge mode classifier, outputting probability distribution of the discharge source belonging to different typical discharge types based on a framework of fusion of a support vector machine and a deep neural network, associating a discharge type label with the highest probability with three-dimensional space coordinates and energy intensity information of the discharge source, and marking the three-dimensional space coordinates and energy intensity information on the partial discharge three-dimensional space distribution image to form an enhanced three-dimensional positioning imaging result with discharge type information.
8. The transformer substation partial discharge positioning and imaging method based on the acoustic phased array is characterized by further comprising the steps of monitoring data dynamic updating and trend analysis, wherein all the steps in the method 1 are repeatedly executed according to a preset period to obtain multi-frame partial discharge three-dimensional space distribution images under different time stamps, a three-dimensional discharge activity database based on time sequences is built, discharge source coordinates, energy intensity and occurrence time in each frame of images are stored in time sequence, space-time clustering analysis is conducted on the three-dimensional discharge activity database, stable discharge hot spots which continuously exist in space or are increased in energy regularity are identified, statistical characteristics of each stable discharge hot spot including average energy increase rate, space position drift rate and discharge frequency change trend are calculated, the development trend of the discharge hot spots is predicted based on the statistical characteristics, and early warning is conducted on hot spots which possibly constitute insulation fault risks in a grading mode.
9. The method for positioning and imaging partial discharge of a transformer substation based on an acoustic phased array according to claim 5, wherein the step of establishing a propagation velocity model of sound waves in a complex environment of the transformer substation specifically comprises: Arranging a plurality of acoustic reference sources with known space coordinates in a transformer substation monitoring area, controlling the acoustic reference sources to emit specific pulse signals, receiving the specific pulse signals through an acoustic phased array sensor array to obtain calibration signal data, inverting and calculating equivalent sound speeds of sound waves on different propagation paths in the monitoring area based on the calibration signal data and the known acoustic reference source coordinates to obtain a group of discrete path sound speed values, performing spatial interpolation on the group of discrete path sound speed values by using a Kriging spatial interpolation algorithm, fitting a sound speed spatial distribution field of sound waves continuously changing in the whole monitoring area, performing associated modeling on the sound speed spatial distribution field and environmental parameters acquired by a temperature sensor and a humidity sensor, establishing a propagation speed model describing dynamic change of sound speeds along with the space position and the environmental parameters, and using the propagation speed model for solving a time difference positioning equation set.
10. The substation partial discharge positioning and imaging method based on the acoustic phased array according to claim 1, wherein the step of deploying the acoustic phased array sensor array specifically comprises: According to the three-dimensional structure model and the electromagnetic interference distribution diagram of a transformer substation monitoring area, an initial point distribution scheme of an acoustic phased array sensor array is planned, all equipment to be monitored is ensured to be covered and strong electromagnetic interference areas are avoided, distributed sensor nodes capable of being flexibly networked are adopted, each node comprises a microphone array module, a signal preprocessing module and a wireless communication module, the sensor nodes are arranged on an insulating support according to the initial point distribution scheme, final accurate three-dimensional coordinates of each array element in each node are measured and recorded to form an accurate geometric position model of the array, and when the system is operated, self-adaptive fine adjustment is carried out through the pointing direction and the gesture of a sensor node of a cloud deck control part according to a primary positioning result and signal quality feedback so as to optimize the receiving sensitivity of discharge signals in a specific direction.

Description

Transformer substation partial discharge positioning and imaging method based on acoustic phased array Technical Field The invention relates to the field of insulation monitoring and fault positioning of power equipment, in particular to a transformer substation partial discharge positioning and imaging method based on an acoustic phased array. Background The safe and stable operation of the power system is a core guarantee of the energy supply reliability, and the transformer substation is used as a key hub for power transmission and conversion, and the insulation performance of the internal power equipment directly influences the overall operation safety of the system. Partial discharge is an important sign of insulation degradation of power equipment, a partial discharge source is accurately positioned in time, the distribution characteristics of the partial discharge source are mastered, accurate basis can be provided for equipment maintenance, and serious power failure accidents caused by insulation fault expansion are avoided. At present, various partial discharge positioning techniques have been developed in the industry, including ultra-high frequency methods, ultrasonic methods, time difference methods, angle measurement methods, and the like. The ultrahigh frequency method is widely applied by virtue of the advantages of strong anti-interference performance and high propagation speed, for example, the prior art CN108490325B discloses a two-section type substation partial discharge signal positioning method, a ultrahigh frequency sensor is used for collecting signals, and the positioning of a discharge point is realized by combining clustering preliminary positioning and compressed sensing accurate positioning, so that the hardware requirement is reduced to a certain extent, and the positioning precision is ensured. However, the prior art still has a plurality of limitations that firstly, the positioning dimension is limited, most of the positioning dimension can only realize two-dimensional plane positioning, three-dimensional space coordinates of a discharge source cannot be obtained, the multi-equipment and multi-level monitoring requirements under the complex transformer substation environment are difficult to meet, secondly, the visual imaging function is lacking, only a single or a plurality of discrete positioning points can not be output, the space distribution situation and the energy intensity difference of partial discharge can not be visually presented, the fault influence range is not easy to be judged by operation staff, thirdly, the multi-discharge source separation capability is insufficient, when a plurality of simultaneous partial discharge sources exist in the transformer substation, the existing method is easy to generate signal confusion, the positioning deviation is increased, fourthly, the anti-interference performance is to be improved, the complex electromagnetic environment and background noise in the transformer substation are easy to interfere the acquired signals, the positioning accuracy is influenced, and fifthly, the multi-dimensional information of the discharge signals is not fully combined, and the comprehensive analysis of the characteristics and the positions of the discharge sources is difficult to realize. These problems lead to the prior art that it is difficult to fully meet the actual requirements of accurate monitoring, visual presentation and efficient operation and maintenance of the partial discharge of the transformer substation, and a partial discharge monitoring technology capable of realizing three-dimensional positioning, visual imaging, multi-source separation and high anti-interference capability is needed. Disclosure of Invention The invention provides a transformer substation partial discharge positioning and imaging method based on an acoustic phased array, and aims to solve the problems of limited partial discharge positioning dimension, lack of a visual imaging function, poor multi-discharge source separation effect and insufficient anti-interference capability in the prior art. In order to achieve the above purpose, the following technical scheme is adopted. A transformer substation partial discharge positioning and imaging method based on an acoustic phased array comprises the following steps of deploying an acoustic phased array sensor array in a transformer substation monitoring area, collecting multichannel original acoustic signals generated by partial discharge to obtain array time domain signal data, preprocessing the array time domain signal data, obtaining preprocessed array frequency domain signal data, calculating a spatial spectrum through a self-adaptive beam forming algorithm based on the preprocessed array frequency domain signal data to obtain initial spatial spectrum distribution containing energy peaks, carrying out peak detection and cluster analysis on the energy peaks in the initial spatial spectrum distribution to sepa