Method for optimizing operation state data of surge arrester based on fuzzy clustering and dynamic weighted fusion

Abstract

The application discloses a kind of based on fuzzy clustering and dynamic weighting fusion's arrester operating state data optimization method, this method includes: step 1: the extraction and normalization processing of arrester full current signal and frequency domain feature;Step 2: setting the density peak method initialization cluster center of integrated decision value condition;Step 3: using Gaussian kernel function to the data after processing Nonlinear mapping, construct the improved fuzzy clustering (IFCM) objective function with regular term and solve;Step 4: combined with fuzzy membership and dynamic weighting fusion mechanism of Mahalanobis distance constructs data comprehensive score model;Step 5: based on adaptive threshold screening abnormal data.The method effectively overcomes the sensitivity of traditional fuzzy clustering algorithm to initial value and the limitation of Euclidean distance, can automatically identify and eliminate strong volatility, low-quality data points disturbed, significantly improve the accuracy and robustness of health assessment model.

Description

Technical Field

[0001] This invention belongs to the field of power equipment condition monitoring and assessment technology, specifically involving a method for optimizing surge arrester operating status data based on fuzzy clustering and dynamic weighted fusion. Background Technology

[0002] Zinc oxide surge arresters are widely used in power transmission and distribution network systems, and their operating status directly affects the system's insulation level and safe and stable operation. Factors affecting the operating status of surge arresters include pollution, ambient temperature, rainfall, temperature rise, phase-to-phase temperature difference, total current, rate of change of total current, resistive leakage current, and rate of change of leakage current. The main challenges currently facing online monitoring of surge arrester operating status are:

[0003] 1. Data source diversity and high interference: The monitoring data comes from multiple sensors such as voltage, temperature, and total current. The data is easily affected by factors such as on-site electromagnetic interference, sensor noise, and instantaneous changes in operating conditions (such as operating overvoltage and lightning strikes), exhibiting strong fluctuations and containing a large number of abnormal pulse points.

[0004] 2. Limitations of traditional FCM: Traditional fuzzy C-means (FCM) clustering is sensitive to the initial cluster centers, and the use of Euclidean distance makes it difficult to handle the nonlinear and non-spherical distribution of surge arrester operating data under complex operating conditions.

[0005] 3. Lack of objective quality quantification standards: The lack of objective and quantitative standards to evaluate and screen the quality of massive monitoring data leads to insufficient reliability of the data input into the health assessment model, affecting the accuracy and credibility of the assessment results.

[0006] Therefore, there is an urgent need to develop an optimization method that can effectively integrate multi-source information, automatically identify and eliminate interfering data, and objectively quantify data quality, so as to provide a high-quality data foundation for accurate condition assessment of surge arresters. Summary of the Invention

[0007] The purpose of this invention is to propose a method for optimizing the operating status data of surge arresters based on fuzzy clustering and dynamic weighted fusion. This method can effectively integrate multi-source information, automatically identify and eliminate interfering data, and objectively quantify data quality, thereby significantly improving the accuracy and robustness of the health assessment model.

[0008] The technical solution to achieve the purpose of this invention is as follows:

[0009] A method for optimizing surge arrester operating status data based on fuzzy clustering and dynamic weighted fusion includes:

[0010] Step S1: Collect raw data of surge arrester operation monitoring, perform frequency domain feature extraction and normalization processing, and construct surge arrester sample set;

[0011] Step S2: Based on the surge arrester sample set, initialize the cluster centers using the density peak method to determine the comprehensive decision value conditions;

[0012] Step S3: Based on the initial cluster centers, a Gaussian kernel function is used to perform nonlinear mapping on the surge arrester sample data to construct a fuzzy clustering objective function with a regularization term;

[0013] Step S4: Solve the fuzzy clustering objective function to obtain the fuzzy membership degree. Combine the dynamic weighted fusion mechanism of fuzzy membership degree and Mahalanobis distance to construct a comprehensive data scoring model and obtain the data scores of all samples.

[0014] Step S5: Remove abnormal samples with scores lower than the adaptive scoring threshold.

[0015] Furthermore, the raw data for monitoring the operation of the surge arrester includes power supply voltage, ambient temperature, humidity, total atmospheric pressure, and raw waveform signal of the full current.

[0016] Furthermore, Fourier transform is used to extract frequency domain features, obtain spectrum information and extract key feature parameters, including the peak value of the total current, the effective value of the total current, the effective value of the fundamental wave, the initial phase angle of the fundamental wave, the amplitude of the third harmonic, the effective value of the third harmonic, the amplitude of the fifth harmonic, and the effective value of the fifth harmonic.

[0017] Further, step S2 specifically includes:

[0018] S21. Calculate the Euclidean distance between any two points in the surge arrester sample set. ,in N is the total number of samples collected;

[0019] S22, Euclidean distance between all sample points Ascending order processing to form a set Determine the cutoff distance , For index value:

[0020] ;

[0021] in, This is a weighting factor for the operation of surge arresters. It is a rounding function;

[0022] S23, Based on Euclidean distance between sample points and cutoff distance Calculate local density minimum distance ;

[0023] ;

[0024] S24. Normalize the local density and minimum distance respectively to obtain... and Calculate the comprehensive decision value for each sample point: ;

[0025] S25. Arrange the comprehensive decision values ​​in descending order to form a sequence. Calculate the adjacent differences and determine the index position C corresponding to the largest difference as the total number of cluster centers. Then, determine all points in the sequence with indices less than C as the initial cluster centers. ,in .

[0026] Furthermore, the fuzzy clustering objective function with regularization in step S3 is:

[0027] ;

[0028] in, For Gaussian kernel function, As cluster center, Let be the membership matrix, indicating that the i-th sample belongs to the k-th cluster center. The degree of fuzziness, where m is the fuzzy weighting index; This is the regularization coefficient.

[0029] Furthermore, the data comprehensive scoring model in step S4 is as follows:

[0030] ;

[0031] in, Representing data points Dynamic weights, The membership matrix, .

[0032] Furthermore, the fuzzy weighting index m=1.

[0033] Furthermore, the dynamic weights for:

[0034] ;

[0035] ;

[0036] in, and They are respectively The mean and covariance; Represents the j-th sample The degree of deviation from the normal statistical pattern, It is a positive number.

[0037] Furthermore, the adaptive scoring threshold is:

[0038]

[0039] in, For adaptive coefficients, The mean of all sample scores. is the standard deviation of all scores.

[0040] Furthermore, the adaptive coefficient The range of values ​​is .

[0041] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0042] 1. Strong noise resistance: The introduction of Gaussian kernel function and regularization term can effectively suppress the impact of pulse interference caused by instantaneous changes in substation operating conditions or electromagnetic interference, thereby improving data quality.

[0043] 2. Dynamic adaptability and high precision: The dynamic weights calculated based on Mahalanobis distance are updated in real time, which can adaptively evaluate the quality deviation of data under different working conditions (such as changes in ambient temperature and humidity), and significantly improve the accuracy of data quality evaluation.

[0044] 3. High computational efficiency: The density peak initialization strategy with set comprehensive decision value conditions avoids random initialization, can converge quickly, and effectively reduce the number of iterations.

[0045] 4. Improved robustness: The optimized data is more suitable for the surge arrester health assessment model, improving the accuracy and reliability of the surge arrester operating status assessment. Attached Figure Description

[0046] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments will be briefly described below. Obviously, the drawings described below only relate to some embodiments of the present invention and are not intended to limit the present invention.

[0047] Figure 1 This is a flowchart of the proposed method for optimizing surge arrester operating status data based on improved fuzzy clustering and dynamic weighted fusion.

[0048] Figure 2 This is a diagram showing the results of the improved fuzzy clustering method in an embodiment of the present invention;

[0049] Figure 3 This is a box plot of the health score indicators before and after optimization in an embodiment of the present invention;

[0050] Figure 4This is a comparison diagram of the total current signal of the surge arrester before and after the method optimization in this embodiment of the invention. Detailed Implementation

[0051] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0052] like Figure 1 As shown, this invention proposes a method for optimizing surge arrester operating status data based on improved fuzzy clustering and dynamic weighted fusion. This method effectively overcomes the limitations of traditional fuzzy clustering algorithms, such as sensitivity to initial values ​​and Euclidean distance. It can automatically identify and remove low-quality data points with strong fluctuations and interference, significantly improving the accuracy and robustness of the health assessment model. Specifically, it includes:

[0053] S1. Collect raw data from surge arrester operation monitoring, and perform frequency domain feature extraction and normalization processing.

[0054] Specifically, this invention collects historical operating data of surge arresters at a monitoring site under different environmental conditions, covering total current, total current change rate, resistive leakage current, leakage current change rate, ambient temperature, phase-to-phase temperature difference, rainfall, pollution level, etc.

[0055] The acquired raw current signal is subjected to Fast Fourier Transform (FFT) to convert the time domain signal into a frequency domain signal. Key feature parameter values ​​are extracted from the FFT spectrum results, including: peak value of total current, effective value of total current, effective value of fundamental wave, initial phase angle of fundamental wave, amplitude of third harmonic, effective value of third harmonic, amplitude of fifth harmonic, and effective value of fifth harmonic.

[0056] All characteristic parameters (including environmental and electrical characteristic parameters) obtained from collection and calculation are normalized to obtain the following: N represents the total number of samples collected; this eliminates the influence of dimensions and prepares for subsequent analysis.

[0057] S2. Initialize cluster centers using the density peak method with comprehensive decision value conditions.

[0058] To overcome the limitation of traditional FCM algorithms being sensitive to initial cluster centers, this invention introduces a method based on sample local density (…). ) and minimum distance ( This paper describes a density peak initialization method for surge arresters. In surge arrester operation status data monitoring, data points with high local density usually represent common or stable operating conditions, while data points in low-density areas may correspond to abnormal or transitional states. By automatically selecting initial cluster centers, the randomness of traditional FCM is avoided, ensuring that the selected center points fall in dense areas of data distribution, making it more likely to approach the global optimum. Compared with traditional FCM, this method improves the algorithm's convergence speed and computational efficiency, enhances the accuracy of identifying typical surge arrester operating modes, and provides a reliable foundation for subsequent state division and fault diagnosis.

[0059] To more accurately select initial cluster centers, this invention, based on the original density peak method, sets a comprehensive decision value condition, which simultaneously considers the relative magnitude of local density and minimum distance. This dual selection mechanism ensures that the selected centers are located in data-dense regions while effectively distinguishing different data clusters, thereby further improving the accuracy and robustness of the clustering algorithm.

[0060] Specifically, the density peak method for initializing cluster centers includes the following steps:

[0061] S21. Calculate the Euclidean distance between any two points in the surge arrester sample set. ,in ;

[0062] S22, Euclidean distance between all sample points Ascending order processing to form a set Determine the cutoff distance , For index value:

[0063]

[0064] in, The weighting factor for the operation of surge arresters is typically 0.01 to 0.02.

[0065] S23, Calculate local density minimum distance ;

[0066]

[0067] S24. Normalize the local density and minimum distance respectively to obtain... and Calculate the comprehensive decision value for each point: ;

[0068] S25. Arrange the comprehensive decision values ​​in descending order to form a sequence. Calculate the adjacent differences and determine the index position C corresponding to the largest difference as the total number of cluster centers. Then, determine all points in the sequence with indices less than C as the initial cluster centers. , .

[0069] S3. Use the Gaussian kernel function to perform nonlinear mapping on the processed data, construct the improved fuzzy clustering (IFCM) objective function with regularization term, and solve it.

[0070] To address the problem of nonlinear data distribution, this invention introduces a Gaussian kernel function to perform nonlinear mapping on the samples, making the data easier to cluster in a high-dimensional feature space. The Gaussian kernel function formula is as follows, where bandwidth is the parameter:

[0071]

[0072] Specifically, the improved fuzzy clustering objective function with regularization in S3 is as follows:

[0073]

[0074] in, For Gaussian kernel function, As cluster center, Let be the membership matrix, indicating that the i-th sample belongs to the k-th cluster center. The degree of fuzziness, where m is the fuzzy weighting index; The regularization coefficient serves as a smoothing factor, and the regularization term is... Used to balance cluster confidence and suppress extreme biases of outlier data points toward fuzzy membership.

[0075] like Figure 2 The diagram shows the clustering results after dimensionality reduction of high-dimensional frequency domain features using the PCA method. Each data point represents a record of surge arrester operation data. Different clusters are identified by the IFCM (Improved Fuzzy C-means) algorithm using color. The cluster boundaries are clear and the cluster centers are evenly distributed.

[0076] The density peak initialization and kernel function mapping strategy proposed in this invention can effectively improve clustering stability and enhance the ability to distinguish outlier data points; points that are on the edge or isolated can be used as candidate outliers to be removed later through a dynamic scoring mechanism.

[0077] S4. A data comprehensive scoring model is constructed by combining the dynamic weighted fusion mechanism of fuzzy membership degree and Mahalanobis distance.

[0078] Specifically, the steps in S4 include:

[0079] S41. Calculate the Mahalanobis distance of each feature dimension of each surge arrester sample set based on the historical normal operating condition data of the surge arrester:

[0080]

[0081] in, and They are respectively The mean and covariance; Represents the i-th sample The degree of deviation from the normal statistical pattern;

[0082] S42. Calculate the dynamic weight of each sample data. :

[0083]

[0084] in, Representative data points The credibility or weight in quality assessment has dynamic adaptability; It is a very small positive number used to ensure numerical stability, and is generally taken to be a value of 1. .

[0085] S43, Fusion Clustering Membership Degree and dynamic weights Construct a comprehensive quality score for the surge arrester sample data. :

[0086]

[0087] like Figure 3 As shown in the figure, the distribution of the two sets of data on the health score index before and after optimization was compared using a box plot. The horizontal axis represents the processing status (before optimization / after optimization), and the vertical axis represents the health score (ranging from 0 to 1). It can be seen from the figure that the overall score of the optimized data shifted upward, with significant improvements in both the mean and median. At the same time, the number of outliers decreased, and the box became more concentrated, indicating that the data quality was improved and the volatility decreased.

[0088] S5. Remove outliers based on adaptive scoring thresholds.

[0089] Specifically, the adaptive scoring threshold is calculated using the mean-standard deviation threshold method. Remove Low-quality data points;

[0090]

[0091] in, This is an adaptive coefficient used to control the stringency of the screening process. The range of values ​​is ; The mean of all scores represents the average data quality level of the current dataset. The standard deviation of all scores reflects the variability in quality; the threshold is... An acceptable minimum quality standard was set; the calculation of adaptive thresholds ensured the objectivity and consistency of data screening, without relying on human experience to set fixed empirical values.

[0092] In surge arrester data monitoring, these abnormal data points are usually data records affected by electromagnetic pulse interference, transient sensor failures, or sudden changes in operating conditions, and their quality scores are significantly lower than those of other data points. Average level. Effectively eliminating highly volatile spikes can significantly improve the stability and continuity of data trends, providing a high-quality and reliable data foundation for subsequent health assessment models.

[0093] Figure 4 The image shows a comparison of the surge arrester's total current signal before and after optimization in this embodiment of the invention. The comparison results show that there are pulse points with strong fluctuations in the original data. After fuzzy comprehensive evaluation, pulse interference data is removed, which can significantly improve the stability and continuity of the data trend. The data after comprehensive evaluation is more suitable for the health assessment model.

[0094] In summary, the innovation of this invention lies in:

[0095] 1. Improved Fuzzy C-means Clustering: Traditional fuzzy C-means (FCM) clustering is sensitive to the initial cluster centers, and the use of Euclidean distance is difficult to handle the nonlinear and non-spherical distribution of surge arrester operating data under complex operating conditions. New strategies need to be introduced to overcome the randomness of traditional FCM and the limitations of distance measurement in order to accurately identify the typical operating modes of surge arresters.

[0096] 2. Dynamic weighted fusion mechanism: Introducing Mahalanobis distance and data noise suppression weights to achieve dynamic and objective quantification of data quality;

[0097] 3. Adaptive scoring and threshold filtering: Automatically determines thresholds based on statistical characteristics to effectively identify and eliminate abnormal fluctuation points.

[0098] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for optimizing surge arrester operating status data based on fuzzy clustering and dynamic weighted fusion, characterized in that, include: Step S1: Collect raw data of surge arrester operation monitoring, perform frequency domain feature extraction and normalization processing, and construct surge arrester sample set; Step S2: Based on the surge arrester sample set, initialize the cluster centers using the density peak method to determine the comprehensive decision value conditions; Step S3: Based on the initial cluster centers, a Gaussian kernel function is used to perform nonlinear mapping on the surge arrester sample data to construct a fuzzy clustering objective function with a regularization term; Step S4: Solve the fuzzy clustering objective function to obtain the fuzzy membership degree. Combine the dynamic weighted fusion mechanism of fuzzy membership degree and Mahalanobis distance to construct a comprehensive data scoring model and obtain the data scores of all samples. Step S5: Remove abnormal samples with scores lower than the adaptive scoring threshold; Step S2 specifically includes: S21. Calculate the Euclidean distance between any two points in the surge arrester sample set. ,in N is the total number of samples collected; S22, Euclidean distance between all sample points Ascending order processing to form a set Determine the cutoff distance , For index value: ； in, This is a weighting factor for the operation of surge arresters. It is a rounding function; S23, Based on Euclidean distance between sample points and cutoff distance Calculate local density minimum distance ; ； S24. Normalize the local density and minimum distance respectively to obtain... and Calculate the comprehensive decision value for each sample point: ; S25. Arrange the comprehensive decision values ​​in descending order to form a sequence. Calculate the adjacent differences and determine the index position C corresponding to the largest difference as the total number of cluster centers. Then, determine all points in the sequence with indices less than C as the initial cluster centers. ,in .

2. The method for optimizing surge arrester operating status data based on fuzzy clustering and dynamic weighted fusion according to claim 1, characterized in that, The raw data for monitoring the operation of the surge arrester includes power supply voltage, ambient temperature, humidity, total atmospheric pressure, and raw waveform signal of the full current.

3. The method for optimizing surge arrester operating status data based on fuzzy clustering and dynamic weighted fusion according to claim 2, characterized in that, Fourier transform is used for frequency domain feature extraction to obtain spectrum information and extract key feature parameters, including the peak value of the total current, the effective value of the total current, the effective value of the fundamental wave, the initial phase angle of the fundamental wave, the amplitude of the third harmonic, the effective value of the third harmonic, the amplitude of the fifth harmonic, and the effective value of the fifth harmonic.

4. The method for optimizing surge arrester operating status data based on fuzzy clustering and dynamic weighted fusion according to claim 1, characterized in that, The fuzzy clustering objective function with regularization in step S3 is: ； in, For Gaussian kernel function, As cluster center, Let be the membership matrix, indicating that the i-th sample belongs to the k-th cluster center. The degree of fuzziness, where m is the fuzzy weighting index; This is the regularization coefficient.

5. The method for optimizing surge arrester operating status data based on fuzzy clustering and dynamic weighted fusion according to claim 4, characterized in that, The data comprehensive scoring model in step S4 is as follows: ； in, Representing data points Dynamic weights, The membership matrix, .

6. The method for optimizing surge arrester operating status data based on fuzzy clustering and dynamic weighted fusion according to claim 5, characterized in that, The fuzzy weighting index m=1.

7. The method for optimizing surge arrester operating status data based on fuzzy clustering and dynamic weighted fusion according to claim 5, characterized in that, The dynamic weight for: ； ； in, and They are respectively The mean and covariance; Represents the j-th sample The degree of deviation from the normal statistical pattern, It is a positive number.

8. The method for optimizing surge arrester operating status data based on fuzzy clustering and dynamic weighted fusion according to claim 1, characterized in that, The adaptive scoring threshold is: ； in, For adaptive coefficients, The mean of all sample scores. is the standard deviation of all scores.

9. The method for optimizing surge arrester operating status data based on fuzzy clustering and dynamic weighted fusion according to claim 8, characterized in that, The adaptive coefficient The range of values ​​is .

Images