A method for locating partial discharge position of complete set of equipment in 10KV transformer area

Through signal acquisition and intelligent algorithms, the local discharge power supply of 10KV transformer is accurately positioned, and the problems of inaccurate positioning and low efficiency in the existing technology are solved, efficient and accurate local discharge detection is achieved, and the operation risks and maintenance costs of power equipment are reduced.

CN118584399BActive Publication Date: 2024-11-05HONGGUANG ELECTRIC GROUP CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411053594.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-02
Publication Date
2024-11-05
Estimated Expiration
2044-08-02

AI Technical Summary

Technical Problem

There are problems such as inaccurate positioning, low detection efficiency and interference from environmental factors in the partial discharge detection of complete sets of equipment in the existing 10KV transformer table, which affects maintenance timeliness and safety.

Method used

The steps of signal acquisition, signal preprocessing, feature extraction, pattern recognition, sound source positioning and data analysis are adopted, combined with the transformer structure and material characteristics, and the local discharge power is accurately positioned through intelligent algorithms and sound source positioning technology, and the results are displayed using visualization technology.

Benefits of technology

It significantly improves detection efficiency and positioning accuracy, reduces misjudgment and misjudgment, reduces maintenance costs and failure rates, and enhances the real-time monitoring capabilities of the power system and the stability and safety of equipment operation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118584399B_ABST
    Figure CN118584399B_ABST
Patent Text Reader

Abstract

The present invention discloses a method for locating the local discharge position of a complete set of equipment in a 10KV transformer substation, and relates to the technical field of discharge detection. The present invention comprises the following steps: signal acquisition: ultrasonic sensors or other types of sensors are arranged at key positions of the transformer to capture signals generated by local discharge; the sensor type is determined according to the characteristics of the transformer and the characteristics of the local discharge signal; the present invention significantly improves the detection efficiency and positioning accuracy of local discharge of complete sets of equipment in a 10KV transformer substation by adopting signal processing technology and intelligent algorithms. Compared with traditional detection methods, it not only reduces the dependence on professional technicians, but also greatly reduces the maintenance cost and equipment failure rate caused by misjudgment or missed judgment. Through automated feature extraction and pattern recognition, a rapid response to local discharge signals is achieved, and the real-time monitoring capability of the power system is enhanced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of discharge detection, and in particular to a method for locating a partial discharge position of a complete set of equipment in a 10KV transformer station area. Background Art

[0002] During the operation of existing 10KV transformer substation equipment, partial discharge is one of the main causes of equipment failure and performance degradation. Partial discharge usually occurs at insulation weaknesses inside the transformer, such as windings, insulating bushings or tap switches. Although the existing technology has adopted a variety of detection methods, these methods often have problems such as inaccurate positioning, low detection efficiency, high cost or sensitivity to environmental conditions. These problems limit the timeliness and effectiveness of transformer maintenance and increase the operation risk of the power system;

[0003] Traditional partial discharge detection methods mainly rely on the measurement of electrical signals, such as identifying the characteristics of partial discharge signals through frequency domain analysis. However, these methods often have difficulty in accurately locating the discharge source, especially in complex transformer structures. In addition, partial discharge signals may be interfered by various factors, such as environmental noise and equipment aging, making signal identification and analysis more difficult.

[0004] Existing technologies often lack sufficient flexibility and accuracy when dealing with these problems, resulting in the inability to timely detect and handle partial discharge events, thereby affecting the reliability and safety of the transformer.

[0005] In order to solve the above defects, a technical solution is now provided. Summary of the invention

[0006] The purpose of the present invention is to solve the problems of inaccurate positioning, low detection efficiency and interference from environmental factors in the existing partial discharge detection of 10KV transformer area complete equipment, and to propose a method for locating the partial discharge position of 10KV transformer area complete equipment.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] A method for locating the partial discharge position of a complete set of equipment in a 10KV transformer station area comprises the following steps:

[0009] S1: Signal acquisition: Sensors are placed at key locations of the transformer to capture partial discharge signals;

[0010] S2: Signal preprocessing: perform preprocessing operations on the collected signals;

[0011] S3: Feature extraction: extracting features for identifying partial discharge from the preprocessed signal;

[0012] S4: Pattern recognition: Use algorithms to classify and identify features and distinguish partial discharge signals;

[0013] S5: Sound source localization: Use sound source localization technology to determine the source of the partial discharge signal;

[0014] S6: Data analysis: Analyze the accuracy of positioning results based on transformer structure and material characteristics;

[0015] S7: Result visualization: Use visualization technology to display the positioning results.

[0016] Furthermore, the specific steps of S1 are as follows:

[0017] Select the corresponding sensor type according to the transformer characteristics and partial discharge signal characteristics; analyze the transformer structure and determine the high-incidence areas of partial discharge, including windings, insulating bushings or tap switches;

[0018] Establish monitoring points in the identified high-incidence areas and deploy sensors, ensuring that sensor coverage and space limitations are taken into account; connect the sensor output signal to the data acquisition system and perform necessary signal conditioning, including amplification and filtering;

[0019] Configure data acquisition system parameters, including sampling rate, resolution, and acquisition time, to capture partial discharge signals; when using multiple sensors, adopt a synchronous acquisition mechanism to ensure consistency in signal acquisition; implement environmental isolation and protection measures.

[0020] Furthermore, the specific operation steps for determining the high-incidence area of ​​partial discharge in S1 are as follows:

[0021] First, the design drawings and technical documents of the transformer are reviewed to obtain the structural distribution, material specifications and insulation system of the transformer, and to identify the key components in the transformer, including windings, insulating bushings, tap changers, cores and oil tanks;

[0022] Analyze the function of each component and the electrical characteristics of all materials, including dielectric constant and withstand voltage rating, and simulate the electric field distribution inside the transformer using computational electromagnetic methods, including finite element analysis;

[0023] Through electric field analysis, find out the area where the local electric field strength is greater than the preset electric field strength and mark it. The marked area is more likely to have partial discharge;

[0024] Evaluate the quality of insulation materials and insulation gaps to identify weaknesses, including the sealing condition of insulation sleeves;

[0025] Analyze mechanical stresses on transformer components, including stresses in windings due to thermal expansion or electromagnetic forces;

[0026] Analyze the impact of environmental factors on transformers, including temperature, humidity and pollutants;

[0027] Then collect and analyze historical fault data of transformers to identify fault occurrence patterns and common areas;

[0028] Identify areas of high incidence of partial discharge.

[0029] Furthermore, the specific operation steps of S2 are as follows:

[0030] Convert the analog signal into a digital signal, select the corresponding sampling rate and quantization accuracy, evaluate the collected signal amplitude, and when it is lower than the preset standard signal amplitude, amplify the signal to improve the signal-to-noise ratio;

[0031] Apply low-pass and high-pass filters to remove high- and low-frequency noise, use band-pass filters to retain signals in a specific frequency band, and remove DC components;

[0032] Use moving average or exponential smoothing methods to smooth signals and reduce random noise;

[0033] Use window functions such as Hamming window or Hanning window to reduce edge effects in signal processing;

[0034] Adaptive filters or wavelet transforms are used to remove noise, detect and process outliers, and synchronize multi-sensor signals through timestamps or phase locking.

[0035] Furthermore, the specific operation steps of S3 are as follows:

[0036] Select frequency characteristics, time domain characteristics and amplitude characteristics as analysis targets. The specific process is as follows:

[0037] Time Domain Analysis: Perform peak detection, energy calculation, duration measurement, and waveform shape analysis to identify signal features associated with partial discharge events;

[0038] Apply fast Fourier transform to identify the main frequency components of the signal and analyze the spectrum characteristics;

[0039] Use short-time Fourier transform and wavelet transform to perform multi-scale time-frequency analysis of signals;

[0040] Calculate the mean, variance, skewness, and kurtosis of the signal to assess its statistical properties;

[0041] Then the correlation between different features and partial discharge events is evaluated and quantified as a percentage, recorded as the correlation, and features with a correlation greater than a preset standard correlation are selected, and redundant features are removed to reduce the dimension of the feature space;

[0042] The selected features are normalized and combined into a feature vector for pattern recognition and signal classification.

[0043] Furthermore, the specific operation steps of S4 are as follows:

[0044] a. Model selection: Select the corresponding machine learning or deep learning model based on the feature type and complexity, including support vector machine, random forest, decision tree, K-nearest neighbor, convolutional neural network (CNN), recurrent neural network, long short-term memory network and deep belief network;

[0045] b. Model training: Use the training set data to train the selected model and adjust model parameters such as learning rate, regularization parameter, and network structure to optimize model performance;

[0046] c. Performance evaluation: Use validation sets or cross-validation to evaluate model performance to prevent overfitting or underfitting;

[0047] d. Generalization ability test: Use the test set data to test the generalization ability of the model, and use accuracy, recall, F1 score and confusion matrix indicators to evaluate the classification effect;

[0048] e. Model optimization: Based on the test results, continuously optimize the model parameters or structure to improve the recognition accuracy and identify the key features that affect the prediction results;

[0049] f. Model interpretation and visualization: Explain the decision-making process of the model and use confusion matrix or ROC curve tools to show model performance;

[0050] g. Model deployment: Deploy the trained and optimized model in actual applications to perform real-time or batch signal recognition tasks.

[0051] Furthermore, the specific operation steps of S5 are as follows:

[0052] Select the signals identified as partial discharge from the partial discharge events confirmed in the pattern recognition stage, and re-extract the features required for positioning, including signal arrival time and frequency features;

[0053] For each partial discharge event, the time difference or phase difference of the signals received by each sensor in the sensor array is calculated;

[0054] Using time difference or phase difference, the sound source position is estimated through geometric relationship, and the sound source coordinates are calculated;

[0055] A system of equations including all sensor data is built to pinpoint the sound source location by minimizing the sum of squared measurement errors;

[0056] Implement the sound source localization algorithm as a computer program to automatically process the signal and output the localization result;

[0057] Analyze the positioning results, verify the accuracy of the positioning algorithm by simulating discharge signals, and perform calibration.

[0058] Furthermore, the specific operation steps of S5 are as follows:

[0059] Evaluate the acoustic velocity, density and elastic modulus material properties of transformer components; create an acoustic model of the transformer to simulate the propagation of sound waves inside the transformer; compare the simulation results of the acoustic model with the actual collected sound wave data to verify the accuracy of sound source location;

[0060] Identify the factors that affect the accuracy of sound source localization, and through comprehensive analysis of these factors, determine whether the sound source localization results need to be adjusted or corrected;

[0061] When the sound source localization results need to be adjusted or corrected, the sensor layout, algorithm parameters or environmental conditions are adjusted, and adjustment strategies are formulated, including sensor re-layout, algorithm optimization or environmental control measures, and adjustments are implemented, including repositioning sensors, adjusting algorithm parameters or improving the test environment;

[0062] After the adjustment, re-conduct the sound source localization test, collect new positioning data, compare the positioning results before and after the adjustment, analyze the effects of the adjustment measures, continuously monitor the sound source localization process, and continuously optimize the comprehensive impact value model and adjustment strategy based on new data and feedback.

[0063] Furthermore, the specific operation steps of determining whether the sound source localization result needs to be adjusted or corrected in S6 are as follows:

[0064] Identify all factors that affect the accuracy of sound source localization, quantify each factor, and collect data corresponding to the factors, including sensor calibration data, environmental condition records, and material properties;

[0065] Establish a mathematical model to describe how each influencing factor affects the accuracy of sound source localization, then calculate the impact of each influencing factor on the sound source localization result separately, integrate the interaction of all influencing factors, and evaluate the overall impact of the influencing factors on the localization accuracy;

[0066] Assign a weight to each influencing factor, and calculate the comprehensive impact value based on the weight and impact degree to quantify the overall impact of all influencing factors on the positioning results;

[0067] A threshold is preset. When the comprehensive impact value is greater than the threshold, it indicates that the accuracy of the positioning result is insufficient and the positioning result needs to be adjusted or corrected.

[0068] Furthermore, the specific steps of S7 are as follows:

[0069] Collect the coordinate data, sensor data and analysis results obtained from sound source positioning, and select corresponding visualization software or tools according to the needs of information display;

[0070] Build a 2D or 3D model of the transformer and map the location data onto the model, designing the layout, color scheme and symbols to show the PD location and characteristics;

[0071] Generate a static image of the transformer model and annotate the discharge location, and create an animation to show the change of discharge events over time;

[0072] Incorporate interactive elements into visualizations, including tooltips, hotspots, or sliders, to allow users to explore different views of the data and incorporate sensor layout, signal strength, and timestamp information.

[0073] Compared with the prior art, the present invention has the following beneficial effects:

[0074] (1) The present invention significantly improves the detection efficiency and positioning accuracy of partial discharge of complete sets of equipment in 10KV transformer substations by adopting signal processing technology and intelligent algorithms. Compared with traditional detection methods, it not only reduces the dependence on professional technicians, but also greatly reduces the maintenance cost and equipment failure rate caused by misjudgment or missed judgment. Through automated feature extraction and pattern recognition, it achieves a rapid response to partial discharge signals and enhances the real-time monitoring capability of the power system.

[0075] (2) The present invention, by using sound source localization technology and comprehensive data analysis, can effectively determine the precise location of partial discharge, providing a reliable basis for transformer maintenance and fault diagnosis. This precise positioning capability helps to take preventive measures in a timely manner to avoid potential failures caused by partial discharge, thereby significantly reducing the risk of power equipment operation and improving the stability and safety of power supply;

[0076] (3) In the present invention, the result visualization technology intuitively displays complex data in the form of graphics or animations, which greatly facilitates the technicians' understanding and analysis of the partial discharge location. The interactive data view and integrated sensor information enable users to explore and analyze data more conveniently, improving the intuitiveness and convenience of operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings;

[0078] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0079] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0080] It should be understood that the terms "include" and "comprising" used in the specification and claims of the present disclosure indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0081] It should also be understood that the terms used in this disclosure are only for the purpose of describing specific embodiments and are not intended to limit the disclosure. As used in this disclosure and claims, the singular forms of "a", "an", and "the" are intended to include the plural forms unless the context clearly indicates otherwise. It should also be further understood that the term "and / or" used in this disclosure and claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.

[0082] like Figure 1 As shown, a method for locating the partial discharge position of a complete set of equipment in a 10KV transformer area includes the following steps:

[0083] Signal acquisition: ultrasonic sensors or other types of sensors are placed at key locations of the transformer to capture signals generated by partial discharge;

[0084] According to the characteristics of the transformer and the partial discharge signal characteristics, determine the type of sensor, including ultrasonic sensor, electrical sensor or vibration sensor; analyze the transformer structure and determine the high-incidence areas where partial discharge may occur, including windings, insulating bushings or tap switches. The specific steps are as follows:

[0085] First, the design drawings and technical documents of the transformer are reviewed to obtain the structural distribution, material specifications and insulation system of the transformer, and the key components in the transformer are determined, including windings, insulating bushings, tap changers, cores and oil tanks; the functions of each component and the electrical characteristics of all materials are analyzed, including dielectric constants and withstand voltage levels, and the electric field distribution inside the transformer is simulated using computational electromagnetic methods, including finite element analysis; through electric field analysis, the areas where the local electric field strength is greater than the preset electric field strength are found and marked, and the marked areas are more prone to partial discharge; the quality of the insulating materials and the insulation gaps are evaluated to determine the existing weaknesses, including the sealing condition of the insulating bushings; the mechanical stresses on the transformer components are analyzed, including the stresses caused by thermal expansion or electromagnetic forces in the windings; the impact of environmental factors on the transformer, including temperature, humidity and pollutants, are analyzed, and then the historical fault data of the transformer is collected and analyzed to identify the fault patterns and common areas; and the high-incidence areas of partial discharge are identified;

[0086] Establish monitoring points in the identified high-incidence areas of partial discharge, and deploy sensors at the monitoring points, and ensure that the sensors are positioned to capture partial discharge signals. During the deployment of sensors, the coverage and space limitations of the sensors must be considered;

[0087] The output signal of the sensor is connected to the data acquisition system, including a data acquisition card or other signal processing equipment, and the output signal of the sensor is conditioned, including amplification and filtering, to adapt to subsequent signal processing and analysis. The output signal of the sensor is expressed as: , is the output signal of the sensor at time t, is the maximum amplitude of the signal, is the signal frequency, measured in Hertz (Hz), which represents the number of times a periodic event occurs per second. is the phase of the signal, usually in radians, describing the initial position of the signal at time t=0, where t is the time variable;

[0088] Then configure the parameters of the data acquisition system, including sampling rate, resolution and acquisition time, to accurately capture the local discharge signal; when there are multiple sensors in use, adopt a synchronous acquisition mechanism for subsequent signal fusion and positioning analysis, and then perform environmental isolation and protection to ensure that the sensors and data acquisition system have good environmental isolation and electromagnetic compatibility to prevent external noise interference.

[0089] Signal preprocessing: preprocess the collected signals to improve signal quality;

[0090] Convert analog signals into digital signals, select the corresponding sampling rate and quantization accuracy, evaluate the amplitude of the collected signal, and amplify the signal when it is lower than the preset standard signal amplitude to improve the signal-to-noise ratio; use a low-pass filter to remove high-frequency noise, use a high-pass filter to remove low-frequency noise, or use a band-pass filter to retain the signal within a specific frequency range, remove the DC component in the signal by using a high-pass filter or subtracting the average value of the signal;

[0091] Use moving average or exponential smoothing methods to smooth the signal, reduce the impact of random noise, scale the signal to a uniform scale range to facilitate subsequent processing and comparison, and then apply window functions such as Hamming window and Hanning window to the signal to reduce the edge effect in the signal processing process; use adaptive filters or wavelet transform methods to remove noise components in the signal, detect abnormal points or mutation points in the signal, and replace or delete them, and achieve signal synchronization of multiple sensors through timestamp or phase locking methods.

[0092] Feature extraction: Extract features that are helpful in identifying partial discharge from the preprocessed signal, such as frequency features, time domain features, amplitude features, etc.

[0093] First, determine the type of extracted features, including frequency features, time domain features and amplitude features, and then perform time domain analysis, frequency domain analysis, time-frequency analysis and statistical analysis. The specific process is as follows:

[0094] Time Domain Analysis: Peak Detection: Identify peaks in the signal that may be associated with partial discharge events; Energy Calculation: Calculate the energy of the signal by squaring the signal and integrating it, using the formula: The signal energy E is expressed as is the output signal of the sensor at time t; It represents a tiny time interval and is a variable in the integral, representing the infinite subdivision of time during the integral process; Duration: measures the duration of a partial discharge event; Waveform shape: analyzes the waveform shape of the signal and extracts features such as rise time and fall time;

[0095] Frequency domain analysis: Fourier transform: Perform fast Fourier transform (FFT) on the signal and convert it to the frequency domain; Frequency components: Identify the main frequency components of the signal, which may be related to partial discharge; Spectral characteristics: Analyze the distribution characteristics of the spectrum, such as peak frequency, bandwidth, etc.

[0096] Time-frequency analysis: Short-time Fourier transform (STFT): Perform short-time Fourier transform on the signal to analyze the frequency characteristics of the signal at different time points; Wavelet transform: Use wavelet transform to perform multi-scale time-frequency analysis on the signal;

[0097] Statistical analysis: mean and variance: calculate the mean and variance of the signal to reflect the central tendency and dispersion of the signal; skewness and kurtosis: analyze the skewness and kurtosis of the signal to understand the distribution shape of the signal;

[0098] Then the correlation between different features and local discharge events is evaluated and quantified in percentage, recorded as correlation. Features with correlation greater than the preset standard correlation are selected, and redundant features are removed to reduce the dimension of the feature space. The extracted features are normalized to have a uniform scale for easy comparison and processing, and multiple features are combined into a new feature vector to provide more comprehensive information.

[0099] Pattern recognition: Use machine learning or deep learning algorithms to classify and identify the extracted features to distinguish between partial discharge signals and other types of signals;

[0100] Depending on the feature type and complexity, choose the corresponding machine learning or deep learning model, including:

[0101] Machine learning models: support vector machine (SVM), random forest, decision tree and K-nearest neighbor (KNN); deep learning models: convolutional neural network (CNN), recurrent neural network (RNN), long short-term memory network (LSTM) and deep belief network (DBN);

[0102] Use the training set data to train the selected model, adjust the model parameters, including learning rate, regularization parameters and network structure, to obtain the best performance, use the validation set or cross-validation to evaluate the performance of the model to ensure that the model is not overfitting or underfitting; use the test set data to test the generalization ability of the model, evaluate the performance of the model on unseen data, and use indicators such as accuracy, recall, F1 score, confusion matrix, etc. to evaluate the classification effect of the model;

[0103] Based on the test results, continuously optimize the model parameters or structure, improve the model's recognition accuracy, analyze the impact of features on the model's prediction results, identify the most important features, and may need to return to the feature extraction step for adjustment; explain the model's decision-making process and understand how the model is classified, which will help improve the model's reliability; use confusion matrices, ROC curves and other visualization tools to display the model's classification effect, intuitively display the model's performance, and deploy the trained model to actual applications for real-time or batch signal recognition.

[0104] Sound source localization: Use sound source localization technology, such as triangulation based on time difference or phase difference, least square method, etc. to determine the source location of the partial discharge signal;

[0105] Select signals from the local discharge events identified in the pattern recognition stage, which have been identified as local discharge signals, and replay these identified local discharge signals from the data record for sound source localization analysis; re-extract features used for localization, including arrival time and frequency features of the signal, and for each local discharge event, calculate the time difference or phase difference of the signals received by each sensor in the sensor array;

[0106] Use triangulation to estimate the location of the sound source based on the time difference or phase difference. The sound source coordinates are calculated through geometric relationships, including the sides and angles of the triangle. Specifically: the sound source location (x, y, z) is calculated through the time difference And the speed of sound c is calculated: , , ,in is the time difference between the signals received by the sensors in the sensor array;

[0107] Use the least squares method to optimize and refine the estimation of the sound source location by establishing a system of equations that includes all sensor data and solves for the sound source location by minimizing the sum of squares of all sensor measurement errors;

[0108] The sound source localization algorithm is implemented as a computer program to automatically process signals and output localization results. The output results of the sound source localization algorithm are analyzed to determine the possible location of the partial discharge signal. The accuracy of the localization algorithm is verified by simulating discharge signals at known locations, and necessary calibration is performed.

[0109] Data analysis: Analyze the positioning results based on the structure and material characteristics of the transformer to verify its accuracy;

[0110] Based on the transformer data collected during signal acquisition and the analysis of the transformer structure, the material properties of each transformer component are evaluated, including the speed of sound, density and elastic modulus, which will affect the propagation and attenuation of sound waves. An acoustic model of the transformer is established to simulate the propagation path of sound waves inside the transformer. The acoustic model is used to simulate how sound waves generated by partial discharge propagate inside the transformer and predict the time or phase difference when they reach the sensor.

[0111] Compare the simulation results with the actual collected sound wave data to verify the accuracy of the sound source location. Combine the sound source location results with the transformer structure analysis and material property evaluation to verify the rationality of the location results. Identify the factors that affect the accuracy of the sound source location. Through comprehensive analysis of these factors, determine whether the sound source location results need to be adjusted or corrected. The specific steps are as follows:

[0112] Identify all influencing factors that affect the accuracy of sound source localization, including sensor accuracy, signal attenuation, multipath effect, environmental noise, sensor layout and sound speed model; quantify each influencing factor and collect data corresponding to the influencing factor, including sensor calibration data, environmental condition records and material properties; establish a mathematical model to describe how each influencing factor affects the accuracy of sound source localization, and then calculate the potential impact of each influencing factor on the sound source localization result separately, comprehensively consider the interaction of all influencing factors, and evaluate the overall impact of the influencing factors on the positioning accuracy; assign weights to each influencing factor to reflect its relative importance to the positioning accuracy, and calculate a comprehensive impact value based on the weight and degree of influence. The comprehensive impact value quantifies the overall impact of all influencing factors on the positioning result;

[0113] A threshold is preset. When the comprehensive impact value is greater than the threshold, it indicates that the accuracy of the positioning result is insufficient and the positioning result needs to be adjusted or corrected. Specifically, the adjustment strategy is formulated by adjusting the sensor layout, algorithm parameters or environmental conditions, including sensor re-layout, algorithm optimization or environmental control measures, and the adjustment is implemented, including re-positioning the sensor, adjusting the algorithm parameters or improving the test environment;

[0114] After the adjustment, re-conduct the sound source localization test, collect new positioning data, compare the positioning results before and after the adjustment, analyze the effects of the adjustment measures, continuously monitor the sound source localization process, and continuously optimize the comprehensive impact value model and adjustment strategy based on new data and feedback.

[0115] Result visualization: Using visualization technology, the positioning results are displayed in the form of graphics or animations to help technicians intuitively understand the location of partial discharge;

[0116] Summarize the coordinate data, sensor data and any relevant analysis results obtained from the sound source location, and select the corresponding visualization software or tools according to the type and complexity of information to be displayed, including MATLAB, Python (matplotlib, seaborn library) and visualization software; Create a 2D or 3D model of the transformer, including all relevant internal structures and components, map the location data to the transformer model, and determine the specific representation of the partial discharge location in the model;

[0117] Design visualization layouts, color schemes, and symbols to clearly demonstrate the location and characteristics of partial discharges, generate static images of the transformer model, annotate the locations of partial discharges on the images, and create animations to show the evolution of partial discharges over time or the sequence of discharge events; add interactive elements to the visualization, including tooltips, clickable hotspots, or sliders, to allow users to explore different views of the data; integrate other relevant information in the visualization, including sensor layout, signal strength, and timestamps.

[0118] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for locating the partial discharge position of a 10KV transformer area complete set of equipment, characterized in that: The following steps are involved: S1: Signal acquisition: Sensors are placed at key locations of the transformer to capture partial discharge signals; S2: Signal preprocessing: perform preprocessing operations on the collected signals; S3: Feature extraction: extracting features for identifying partial discharge from the preprocessed signal; S4: Pattern recognition: Use algorithms to classify and identify features and distinguish partial discharge signals; S5: Sound source localization: Use sound source localization technology to determine the source of the partial discharge signal. The specific process is as follows: Select the signals identified as partial discharge from the partial discharge events confirmed in the pattern recognition stage, and re-extract the features required for positioning, including signal arrival time and frequency features; For each partial discharge event, the time difference or phase difference of the signals received by each sensor in the sensor array is calculated; Using time difference or phase difference, the sound source position is estimated through geometric relationship, and the sound source coordinates are calculated; A system of equations including all sensor data is built to pinpoint the sound source location by minimizing the sum of squared measurement errors; Implement the sound source localization algorithm as a computer program to automatically process the signal and output the localization result; Analyze the positioning results, verify the accuracy of the positioning algorithm by simulating discharge signals, and perform calibration; S6: Data analysis: Analyze the accuracy of the positioning results based on the transformer structure and material characteristics. The specific process is as follows: Evaluate the acoustic velocity, density and elastic modulus material properties of transformer components; create an acoustic model of the transformer to simulate the propagation of sound waves inside the transformer; compare the simulation results of the acoustic model with the actual collected sound wave data to verify the accuracy of sound source location; Identify the factors that affect the accuracy of sound source localization, and through comprehensive analysis of these factors, determine whether the sound source localization results need to be adjusted or corrected; When the sound source localization results need to be adjusted or corrected, the sensor layout, algorithm parameters or environmental conditions are adjusted, and adjustment strategies are formulated, including sensor re-layout, algorithm optimization or environmental control measures, and adjustments are implemented, including repositioning sensors, adjusting algorithm parameters or improving the test environment; After the adjustment, re-test the sound source localization, collect new localization data, compare the localization results before and after the adjustment, analyze the effect of the adjustment measures, continuously monitor the sound source localization process, and continuously optimize the comprehensive impact value model and adjustment strategy based on new data and feedback; S7: Result visualization: Use visualization technology to display the positioning results.

2. A method for locating partial discharge position of a 10KV transformer area complete set of equipment according to claim 1, characterized in that: The specific steps of S1 are as follows: Select the corresponding sensor type according to the transformer characteristics and partial discharge signal characteristics; analyze the transformer structure and determine the high-incidence areas of partial discharge, including windings, insulating bushings or tap switches; Establish surveillance sites in identified high-incidence areas and deploy sensors, ensuring that sensor coverage and space limitations are taken into account; Connect the sensor output signal to the data acquisition system and perform signal conditioning, including amplification and filtering; Configure data acquisition system parameters, including sampling rate, resolution, and acquisition time, to capture partial discharge signals; when using multiple sensors, adopt a synchronous acquisition mechanism to ensure consistency in signal acquisition; implement environmental isolation and protection measures.

3. The method for locating the partial discharge position of a 10KV transformer area complete set of equipment according to claim 2, characterized in that: The specific operation steps for determining the high-incidence area of ​​partial discharge in S1 are as follows: First, the design drawings and technical documents of the transformer are reviewed to obtain the structural distribution, material specifications and insulation system of the transformer, and to identify the key components in the transformer, including windings, insulating bushings, tap changers, cores and oil tanks; Analyze the function of each component and the electrical characteristics of all materials, including dielectric constant and withstand voltage rating, and simulate the electric field distribution inside the transformer using computational electromagnetic methods, including finite element analysis; Through electric field analysis, find out the area where the local electric field strength is greater than the preset electric field strength and mark it. The marked area is more likely to have partial discharge; Evaluate the quality of insulation materials and insulation gaps to identify weaknesses, including the sealing condition of insulation sleeves; Analyze mechanical stresses on transformer components, including stresses in windings due to thermal expansion or electromagnetic forces; Analyze the impact of environmental factors on transformers, including temperature, humidity and pollutants; Then collect and analyze historical fault data of transformers to identify fault occurrence patterns and common areas; Identify areas of high incidence of partial discharge.

4. A method for locating partial discharge position of a 10KV transformer area complete set of equipment according to claim 1, characterized in that: The specific operation steps of S2 are as follows: Convert the analog signal into a digital signal, select the corresponding sampling rate and quantization accuracy, evaluate the collected signal amplitude, and when it is lower than the preset standard signal amplitude, amplify the signal to improve the signal-to-noise ratio; Apply low-pass and high-pass filters to remove high- and low-frequency noise, and use band-pass filters to retain signals within a preset frequency range and remove DC components; Use moving average or exponential smoothing methods to smooth signals and reduce random noise; Use Hamming window or Hanning window function to reduce the edge effect of signal processing; Adaptive filters or wavelet transforms are used to remove noise, detect and process outliers, and synchronize multi-sensor signals through timestamps or phase locking.

5. The method for locating the partial discharge position of a 10KV transformer area complete set of equipment according to claim 1, characterized in that: The specific operation steps of S3 are as follows: Select frequency characteristics, time domain characteristics and amplitude characteristics as analysis targets. The specific process is as follows: Time Domain Analysis: Perform peak detection, energy calculation, duration measurement, and waveform shape analysis to identify signal features associated with partial discharge events; Apply fast Fourier transform to identify the main frequency components of the signal and analyze the spectrum characteristics; Use short-time Fourier transform and wavelet transform to perform multi-scale time-frequency analysis of signals; Calculate the mean, variance, skewness, and kurtosis of the signal to assess its statistical properties; Then the correlation between different features and partial discharge events is evaluated and quantified as a percentage, recorded as the correlation, and features with a correlation greater than a preset standard correlation are selected, and redundant features are removed to reduce the dimension of the feature space; The selected features are normalized and combined into a feature vector for pattern recognition and signal classification.

6. A method for locating partial discharge position of a 10KV transformer area complete set of equipment according to claim 1, characterized in that: The specific operation steps of S4 are as follows: a. Model selection: Select the corresponding machine learning or deep learning model based on the feature type and complexity, including support vector machine, random forest, decision tree, K-nearest neighbor, convolutional neural network (CNN), recurrent neural network, long short-term memory network and deep belief network; b. Model training: Use the training set data to train the selected model, adjust the learning rate, regularization parameters, and network structure model parameters to optimize model performance; c. Performance evaluation: Use validation sets or cross-validation to evaluate model performance to prevent overfitting or underfitting; d. Generalization ability test: Use the test set data to test the generalization ability of the model, and use accuracy, recall, F1 score and confusion matrix indicators to evaluate the classification effect; e. Model optimization: Based on the test results, continuously optimize the model parameters or structure to improve the recognition accuracy and identify the key features that affect the prediction results; f. Model interpretation and visualization: Explain the decision-making process of the model and use confusion matrix or ROC curve tools to show model performance; g. Model deployment: Deploy the trained and optimized model in actual applications to perform real-time or batch signal recognition tasks.

7. A method for locating partial discharge position of a 10KV transformer area complete set of equipment according to claim 1, characterized in that: The specific operation steps for determining whether the sound source localization result needs to be adjusted or corrected in S6 are as follows: Identify all factors that affect the accuracy of sound source localization, quantify each factor, and collect data corresponding to the factors, including sensor calibration data, environmental condition records, and material properties; Establish a mathematical model to describe how each influencing factor affects the accuracy of sound source localization, then calculate the impact of each influencing factor on the sound source localization result separately, integrate the interaction of all influencing factors, and evaluate the overall impact of the influencing factors on the localization accuracy; Assign a weight to each influencing factor, and calculate the comprehensive impact value based on the weight and impact degree to quantify the overall impact of all influencing factors on the positioning results; A threshold is preset. When the comprehensive impact value is greater than the threshold, it indicates that the accuracy of the positioning result is insufficient and the positioning result needs to be adjusted or corrected.

8. The method for locating the partial discharge position of a 10KV transformer area complete set of equipment according to claim 1, characterized in that: The specific steps of S7 are as follows: Collect the coordinate data, sensor data and analysis results obtained from sound source positioning, and select corresponding visualization software or tools according to the needs of information display; Build a 2D or 3D model of the transformer and map the location data onto the model, designing the layout, color scheme and symbols to show the PD location and characteristics; Generate a static image of the transformer model and annotate the discharge location, and create an animation to show the change of discharge events over time; Incorporate interactive elements into visualizations, including tooltips, hotspots, or sliders, to allow users to explore different views of the data and incorporate sensor layout, signal strength, and timestamp information.