A method for multi-parameter collaborative monitoring of a sulfur hexafluoride insulated device

Abstract

The application discloses a kind of multi-parameter collaborative monitoring methods of sulfur hexafluoride insulation equipment, it is related to the technical field of equipment monitoring, original parameter is collected, the original parameter includes sulfur hexafluoride concentration value, sulfur hexafluoride density value and original partial discharge signal;Discharge pretreatment is carried out to original partial discharge signal, and local discharge signal is obtained;Point spread function matrix is constructed, and local discharge signal is sparsely reconstructed in combination with orthogonal matching pursuit algorithm, and the coordinate and discharge quantity of local discharge source are obtained;Threshold value is judged based on sulfur hexafluoride concentration value, sulfur hexafluoride density value, the coordinate and discharge quantity of local discharge source, and hierarchical early warning result is obtained.The early discovery ability and fault positioning accuracy of large sulfur hexafluoride insulation equipment insulation defect are significantly improved, and reliable technical support is provided for equipment state evaluation and intelligent operation and maintenance.

Description

Technical Field

[0001] This invention relates to the technical field of equipment monitoring, and in particular to a multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment. Background Technology

[0002] In recent years, sulfur hexafluoride (SF6) insulating gas has been widely used in large-scale power equipment such as GIS switchgear due to its excellent insulation properties. These large-scale devices have complex internal spaces and long electrical distances, meaning any local insulation defect can develop into a through-type fault. Traditional single-parameter monitoring methods have significant limitations, especially for large equipment. When partial discharge occurs, they can only determine whether a discharge exists, but cannot accurately locate the defect within the vast equipment space, greatly hindering fault diagnosis and maintenance. Furthermore, discharge signal extraction is poor in environments with strong electromagnetic interference, and the ability to identify multiple discharge sources is insufficient, making it impossible to accurately assess the severity of insulation degradation. Therefore, there is an urgent need for a collaborative monitoring method that can integrate multi-parameter information to accurately locate internal defects in large-scale equipment, providing a reliable basis for equipment condition assessment and fault early warning.

[0003] Currently, Chinese patent application number CN202310313307.6 discloses a method and apparatus for monitoring and fault analysis of gas-insulated equipment. This invention relates to the field of electrical equipment condition monitoring technology, and is a method and apparatus for monitoring and fault analysis of gas-insulated equipment. The former is performed according to the following steps: First, before the gas-insulated equipment is put into operation, a required amount of trifluoroiodomethane is injected into the sulfur hexafluoride gas in each chamber of the gas-insulated equipment; Second, when the gas-insulated equipment is running or shut down due to a fault, the water content and iodide ion content in the sulfur hexafluoride gas in each chamber are monitored, and the operating status of the gas-insulated equipment is evaluated and fault analyzed based on the water content and iodide ion content of the sulfur hexafluoride gas. This invention avoids the significant discrepancy between the detection data of the gas-insulated equipment's operating status and its shutdown status, improves monitoring accuracy, monitors the actual operating status of the equipment online, and provides a reliable evaluation basis for the operating status and fault analysis of gas-insulated equipment. However, the related technologies cannot use the improved orthogonal matching pursuit algorithm to locate partial discharge signals and solve the problems of inaccurate positioning and unclear quantity of sulfur hexafluoride insulation equipment. The related technologies cannot use signal preprocessing to denoise the original partial discharge signal and suppress mode mixing and boundary effects. Summary of the Invention

[0004] The technical problem solved by this invention is that related technologies cannot use the improved orthogonal matching tracking algorithm to locate partial discharge signals, thus failing to solve the problems of inaccurate positioning and unclear quantity of sulfur hexafluoride insulation equipment. Furthermore, related technologies cannot use signal preprocessing to denoise the original partial discharge signal and suppress mode mixing and boundary effects.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] A multi-parameter coordinated monitoring method for sulfur hexafluoride insulation equipment includes the following steps:

[0007] Step S1: Collect raw parameters;

[0008] The original parameters include sulfur hexafluoride concentration value, sulfur hexafluoride density value, and original partial discharge signal;

[0009] Step S2: Perform discharge preprocessing on the original partial discharge signal to obtain a partial discharge signal;

[0010] Step S3: Construct the point spread function matrix and combine it with the orthogonal matching pursuit algorithm to sparsely reconstruct the partial discharge signal to obtain the coordinates of the partial discharge source and the number of discharges.

[0011] Step S4: Based on the sulfur hexafluoride concentration value, sulfur hexafluoride density value, coordinates of the partial discharge source, and discharge quantity, a threshold judgment is made to obtain a graded early warning result.

[0012] As a preferred embodiment of the multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment described in this invention, step S1 specifically includes:

[0013] Step S11: Install a sulfur hexafluoride sensor and a density relay in the equipment components to monitor the sulfur hexafluoride concentration and density values. The equipment components include the bottom of the equipment, a flange sealing surface, a valve operating mechanism, and a density relay interface.

[0014] Step S12: Set up a sensor array to receive partial discharge signals. The sensor array is a UHF sensor arranged in a rectangular pattern.

[0015] As a preferred embodiment of the multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment described in this invention, step S2 specifically includes:

[0016] Step S21: Add Gaussian white noise to the original partial discharge signal to form a noisy partial discharge signal;

[0017] Step S22: Perform signal segmentation processing on the noisy partial discharge signal to obtain the IMF component and residual term;

[0018] Step S23: Calculate the average value of each IMF component to obtain the IMF component. The calculation formula is as follows:

[0019] ;

[0020] Among them, c i (t) represents the IMF component. Let represent the i-th layer IMF component obtained from the j-th decomposition, and Ne be the overall degree parameter. The partial discharge signal is obtained by adding the IMF component to the residual term obtained from the last EMD decomposition.

[0021] As a preferred embodiment of the multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment described in this invention, the noisy partial discharge signal is subjected to signal segmentation processing, which specifically includes:

[0022] The signal is divided into segments according to a preset segment length M from the starting point of the noisy partial discharge signal. The first segment is extended by m1 sampling points, and the subsequent segments are extended by m2 sampling points before and after each segment.

[0023] The signal is segmented and converted into a matrix form. Each row of the matrix is ​​transformed from a segment of a noisy partial discharge signal. Each element in the matrix represents a sampled value of the noisy partial discharge signal. Each element in the matrix is ​​regarded as a key-value pair, where the key is the position of the sampling point in the noisy partial discharge signal and the value is the sampled value corresponding to the sampling point.

[0024] For each segment of the noisy partial discharge signal, the extreme values ​​of the resulting matrix are determined point by point using the following method:

[0025] If the value of a sampling point is greater than the values ​​of its two adjacent points, it is considered a maximum point.

[0026] If the value of a sampling point is less than the values ​​of its two adjacent points, it is considered a local minimum point.

[0027] The extreme points are sorted out, and the maximum and minimum value matrices are output. The upper and lower envelopes of each signal segment are calculated by cubic spline interpolation to obtain the segmented envelopes. The segmented envelopes are clipped according to the boundary positions of the preset segment length. The clipped segmented envelopes are sorted and connected by keys to output the complete envelope. The mean of the complete envelope is calculated.

[0028] The noisy partial discharge signal is iteratively decomposed based on the envelope mean to extract the IMF components and residual terms of each order of the noisy partial discharge signal.

[0029] As a preferred embodiment of the multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment described in this invention, step S3 specifically includes:

[0030] Step S31: Based on the propagation characteristics of partial discharge signals in sulfur hexafluoride, establish the transfer matrix between the scanning plane and the sensor array;

[0031] The scanning plane is a virtual two-dimensional plane defined around the target sulfur hexafluoride insulating device. The virtual two-dimensional plane is discretized into grid points, and each grid point represents a potential discharge source location.

[0032] Step S32: Construct a point spread function matrix based on the transfer matrix;

[0033] Step S33: The partial discharge signal is sparsely reconstructed using an improved orthogonal matching pursuit algorithm, and the coordinates and residual vector of the partial discharge source are obtained through iteration.

[0034] Step S34: Perform L2 operation on the residual vector to obtain the L2 norm of the residual vector. By monitoring the trend of the L2 norm of the residual vector with the number of iterations, the residual decrease rate is obtained. When the residual decrease rate is lower than a preset threshold, the iteration is terminated to obtain the discharge quantity of the partial discharge power source.

[0035] As a preferred embodiment of the multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment according to the present invention, the establishment of the transfer matrix between the scanning plane and the sensor array specifically includes:

[0036] The propagation delay of the partial discharge signal from the scanning grid points to each UHF sensor is calculated, and a transfer matrix is ​​constructed. The elements of the transfer matrix are represented as follows:

[0037] ;

[0038] Among them, g mn Let r be the element in the m-th row and n-th column of the transition matrix. mn Let ω be the distance from the m-th sensor to the n-th grid point, ω be the angular frequency, and τ be the distance from the m-th sensor to the n-th grid point. mn This corresponds to the propagation delay.

[0039] As a preferred embodiment of the multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment according to the present invention, the construction of the point diffusion function matrix of the transfer matrix specifically includes:

[0040] The point spread function vector at the center of the scanning plane is calculated based on the principle of spatial translation invariance. Corresponding point spread function values ​​are then assigned to other scanning grid points through cyclic displacement, constructing a point spread function matrix whose elements are represented as follows:

[0041] ;

[0042] Among them, A ij w represents the element in the i-th row and j-th column of the point spread function matrix. i Let g be the weight vector of the i-th grid point. j Let H be the array guide vector for the j-th grid point, and the superscript H indicates the conjugate transpose operation.

[0043] As a preferred embodiment of the multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment described in this invention, the improved orthogonal matching pursuit algorithm specifically includes:

[0044] Set the initial iteration number i=0, use the original partial discharge signal as the initial residual vector, set the initial index action set to an empty set, and solve for the index of the maximum position of the inner product between the column vector of the point spread function matrix and the residual vector. The calculation expression is:

[0045] ;

[0046] Where t is the maximum position index, A H e is the conjugate transpose of the point spread function matrix. (i) Let be the residual vector after the i-th iteration;

[0047] Add the maximum position index selected in the i-th iteration to the subscript action set of the (i-1)-th iteration, calculate the generalized inverse matrix of the point spread function matrix corresponding to the subscript action set as the orthogonal projection of the point spread function matrix, solve for the new residual term vector, and calculate the new sparse approximation solution:

[0048] ;

[0049] in Let T(i) be the sparse approximation solution computed after the i-th iteration, and let A be the set of all selected scan grid point indices up to the i-th iteration. T(i) Let T(i) be the submatrix formed by the column vectors specified by T(i). For A T(i) The generalized inverse matrix, e (i-1) This is the residual vector after the (i-1)th iteration;

[0050] Let k be the preset number of iterations. When i ≤ k, the steps are repeated starting from the index of the maximum value of the inner product of the column vector and the residual vector of the solution point diffusion function matrix. When i > k, the iteration stops.

[0051] As a preferred embodiment of the multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment described in this invention, step S4 specifically includes:

[0052] Step S41: Based on the sulfur hexafluoride concentration value, sulfur hexafluoride density value, coordinates of the partial discharge source, and discharge quantity, the monitoring results are obtained, including discharge intensity, defect location, and leakage severity.

[0053] The calculation of the severity of the leak specifically includes:

[0054] The concentration and density values ​​of sulfur hexafluoride were calculated using a piecewise linear function to obtain the concentration score and density score. The severity of the leak was then calculated by weighting the concentration score and density score.

[0055] Step S42: Determine the threshold based on the monitoring results to obtain the graded early warning results.

[0056] As a preferred embodiment of the multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment according to the present invention, the step of threshold judgment of the monitoring results to obtain graded early warning results specifically includes:

[0057] The tiered early warning results include a first early warning, a second early warning, a third early warning, and a fourth early warning;

[0058] When the leakage severity is less than the first threshold and no partial discharge signal is detected, no warning is triggered;

[0059] When the severity of the leakage is less than the first threshold, but a partial discharge signal is detected and the discharge intensity is less than the discharge threshold, the first warning is triggered.

[0060] When the severity of the leakage is less than the first threshold, but a partial discharge signal is detected and the discharge intensity is greater than the discharge threshold, a second warning is triggered.

[0061] When the leakage severity is greater than the first threshold and less than the second threshold, and no partial discharge signal is detected, a second warning is triggered;

[0062] When the leakage severity is greater than the first threshold and less than the second threshold, a partial discharge signal is detected, and the discharge intensity is less than the discharge threshold, a third warning is triggered.

[0063] When the severity of the leakage is greater than the first threshold and less than the second threshold, a partial discharge signal is detected, and the discharge intensity is greater than the discharge threshold, a fourth warning is triggered.

[0064] If the severity of the leak exceeds the second threshold, a fourth warning is triggered.

[0065] The first threshold is less than the second threshold.

[0066] The beneficial effects of this invention are as follows: By optimizing the signal preprocessing workflow, the ability to extract true discharge characteristics from noisy environments is significantly enhanced. By combining the orthogonal matching pursuit algorithm, point spread function matrix, and residual descent rate, sparse reconstruction theory is used to achieve spatial localization and quantity determination of partial discharge sources, effectively avoiding missed and false detections. By establishing a multi-parameter hierarchical early warning model based on leakage severity, discharge intensity, and discharge quantity, differentiated and intelligent early warnings are achieved, from routine monitoring to emergency shutdowns. This provides precise basis for operation and maintenance decisions, realizing a shift from preventative maintenance to predictive precision maintenance, and greatly improving the safe operation level and operation and maintenance efficiency of power equipment. Attached Figure Description

[0067] Figure 1 This is a basic flowchart illustrating a multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment, provided as an embodiment of the present invention. Detailed Implementation

[0068] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0069] Example, refer to Figure 1 As an embodiment of the present invention, a multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment is provided, comprising the following steps:

[0070] Step S1: Collect raw parameters;

[0071] The raw parameters include sulfur hexafluoride concentration, sulfur hexafluoride density, and raw partial discharge signal;

[0072] Step S2: Perform discharge preprocessing on the original partial discharge signal to obtain a partial discharge signal;

[0073] Step S3: Construct the point spread function matrix and combine it with the orthogonal matching pursuit algorithm to sparsely reconstruct the partial discharge signal to obtain the coordinates of the partial discharge source and the number of discharges.

[0074] Step S4: Based on the sulfur hexafluoride concentration value, sulfur hexafluoride density value, coordinates of the partial discharge source, and discharge quantity, a threshold judgment is made to obtain a graded early warning result.

[0075] This invention enhances the ability to extract true discharge characteristics from noisy environments through signal preprocessing. By combining orthogonal matching pursuit algorithm, point spread function matrix, and residual descent rate, it utilizes sparse reconstruction theory to achieve spatial localization and quantity determination of partial discharge sources, effectively avoiding missed and false detections. By establishing a multi-parameter hierarchical early warning system based on leakage severity, discharge intensity, and discharge quantity, it achieves differentiated and intelligent early warning from routine monitoring to emergency shutdowns, providing precise basis for operation and maintenance decisions. This realizes the transformation from preventative maintenance to predictive precision maintenance, greatly improving the safe operation level and operation and maintenance efficiency of power equipment.

[0076] Step S1 specifically includes:

[0077] Step S11: Install a sulfur hexafluoride sensor and a density relay in the equipment components to monitor the sulfur hexafluoride concentration and density values. The equipment components include the bottom of the equipment, the flange sealing surface, the valve operating mechanism, and the density relay interface.

[0078] Step S12: Set the sensor array to receive partial discharge signals. The sensor array consists of UHF sensors arranged in a rectangular pattern.

[0079] In a specific embodiment, dedicated sulfur hexafluoride (SF6) sensors and density relays are installed at potential leakage points and condition monitoring points of the SF6 insulation equipment to acquire real-time concentration and density values ​​reflecting the insulation gas status of the equipment, providing environmental parameter support for condition assessment. Furthermore, a reliable signal acquisition system is constructed by arranging ultra-high frequency (UHF) sensors in a rectangular array that effectively covers spatial signals. This system accurately receives and locates partial discharge signals excited by internal insulation defects, thus providing high-quality data input for subsequent signal processing and feature extraction.

[0080] Step S2 specifically includes:

[0081] Step S21: Add Gaussian white noise to the original partial discharge signal to form a noisy partial discharge signal;

[0082] Step S22: Perform signal segmentation processing on the noisy partial discharge signal to obtain the IMF component and residual term;

[0083] Step S23: Calculate the average value of each IMF component to obtain the IMF component. The calculation formula is as follows:

[0084] ;

[0085] Among them, c i (t) represents the IMF component. Let represent the i-th layer IMF component obtained from the j-th decomposition, and Ne be the overall degree parameter. The partial discharge signal is obtained by adding the IMF component to the residual term obtained from the last EMD decomposition.

[0086] In a specific embodiment, the ensemble empirical mode decomposition (EEMD) algorithm is used to address the mode aliasing problem in partial discharge signal analysis. Gaussian white noise is added to the original signal multiple times to create a noisy partial discharge signal, utilizing the uniform distribution of noise to provide comprehensive extreme point references. Subsequently, each noisy signal undergoes empirical mode decomposition to obtain IMF components and residuals. Finally, by performing ensemble averaging on the intrinsic mode components of the same order in all decomposition results, the randomness and mode aliasing caused by single decomposition are effectively suppressed, thereby extracting mode components with clear physical meaning and stability. Simultaneously, by adding the final intrinsic mode components to the residual term, the integrity of the original signal is ensured, achieving robust and accurate decomposition of the partial discharge signal, laying a solid foundation for subsequent feature extraction and fault diagnosis.

[0087] The noisy partial discharge signal is segmented, and the signal segmentation process specifically includes:

[0088] The signal is divided into segments according to a preset segment length M from the starting point of the noisy partial discharge signal. The first segment is extended by m1 sampling points, and the subsequent segments are extended by m2 sampling points before and after each segment.

[0089] The signal is segmented and converted into a matrix form. Each row of the matrix is ​​transformed from a segment of a noisy partial discharge signal. Each element in the matrix represents a sampled value of the noisy partial discharge signal. Each element in the matrix is ​​regarded as a key-value pair, where the key is the position of the sampling point in the noisy partial discharge signal and the value is the sampled value corresponding to the sampling point.

[0090] For each segment of the noisy partial discharge signal, the extreme values ​​of the resulting matrix are determined point by point using the following method:

[0091] If the value of a sampling point is greater than the values ​​of its two adjacent points, it is considered a maximum point.

[0092] If the value of a sampling point is less than the values ​​of its two adjacent points, it is considered a local minimum point.

[0093] The extreme points are sorted out, and the maximum and minimum value matrices are output. The upper and lower envelopes of each signal segment are calculated by cubic spline interpolation to obtain the segmented envelopes. The segmented envelopes are clipped according to the boundary positions of the preset segment length. The clipped segmented envelopes are sorted and connected by keys to output the complete envelope. The mean of the complete envelope is calculated.

[0094] The noisy partial discharge signal is iteratively decomposed based on the envelope mean to extract the IMF components and residual terms of each order of the noisy partial discharge signal.

[0095] In a specific embodiment, the mean value of the envelope is subtracted from the noisy partial discharge signal to obtain a new signal. This process is repeated until the conditions of IMF are met (i.e., the number of extreme points and the number of zero crossings do not differ by more than 1, and the mean value of the complete envelope is close to zero), and the IMF component is obtained.

[0096] Traditional ensemble empirical mode decomposition (EEMD) suffers from severe boundary effects at the start and end points of long signals due to a lack of extrema. This method addresses this by dividing the long signal into shorter segments of a preset length M and extending these segments before and after the boundaries, transforming the global boundary problem into a local, controllable segment boundary problem. Dividing the signal into smaller segments reduces the amount of data and computational complexity of a single envelope fitting. After calculating the segmented envelope, it is strictly pruned according to the original segment length M, retaining only the core portion unaffected by boundary effects. The resulting complete envelope minimizes global boundary effects, resulting in a more accurate and smoother envelope.

[0097] Step S3 specifically includes:

[0098] Step S31: Based on the propagation characteristics of partial discharge signals in sulfur hexafluoride, establish the transfer matrix between the scanning plane and the sensor array;

[0099] In a specific embodiment, the propagation time delay required for the partial discharge signal to propagate from any scanning grid point on the scanning plane to each sensor in the array can be calculated by using the finite propagation speed of the partial discharge signal in sulfur hexafluoride.

[0100] The scanning plane is a virtual two-dimensional plane defined around the target sulfur hexafluoride insulating device. The virtual two-dimensional plane is discretized into grid points, each grid point representing a potential discharge source location.

[0101] Step S32: Construct the point spread function matrix based on the transition matrix;

[0102] Step S33: The partial discharge signal is sparsely reconstructed using an improved orthogonal matching pursuit algorithm, and the coordinates and residual vector of the partial discharge source are obtained through iteration.

[0103] Step S34: Perform L2 operation on the residual vector to obtain the L2 norm of the residual vector. By monitoring the trend of the L2 norm of the residual vector with the number of iterations, the residual decrease rate is obtained. When the residual decrease rate is lower than a preset threshold, the iteration is terminated to obtain the discharge quantity of the partial discharge power source.

[0104] In a specific embodiment, the preset threshold is the residual decrease rate threshold, with a value ranging from 0.01 to 0.05. When the residual decrease rate during the iteration process is lower than the preset threshold, it indicates that continuing the iteration can no longer significantly improve the signal reconstruction accuracy, so the process terminates, and the number of currently selected grid points is determined as the number of discharges of the partial discharge power supply.

[0105] Establishing the transfer matrix between the scanning plane and the sensor array specifically includes:

[0106] Calculate the propagation delay of the partial discharge signal from the scan grid points to each UHF sensor, and construct the transfer matrix. The elements of the transfer matrix are represented as follows:

[0107] ;

[0108] Among them, g mn Let r be the element in the m-th row and n-th column of the transition matrix. mn Let ω be the distance from the m-th sensor to the n-th grid point, ω be the angular frequency, and τ be the distance from the m-th sensor to the n-th grid point. mn This corresponds to the propagation delay.

[0109] In a specific embodiment, by establishing a transfer matrix between the scanning plane and the sensor array, the physical process of electromagnetic wave propagation of partial discharge signals is transformed into a calculable mathematical relationship. By calculating the propagation delay from each scanning grid point to each sensor and constructing the transfer matrix, the propagation characteristics of the signal from any point in space to the sensor are accurately described.

[0110] The construction of the point spread function matrix from the transition matrix specifically includes:

[0111] The point spread function vector at the center of the scanning plane is calculated based on the principle of spatial translation invariance. Corresponding point spread function values ​​are then assigned to other scanning grid points through cyclic displacement, constructing a point spread function matrix whose elements are represented as follows:

[0112] ;

[0113] Among them, A ij w represents the element in the i-th row and j-th column of the point spread function matrix. i Let g be the weight vector of the i-th grid point. j Let H be the array guide vector for the j-th grid point, and the superscript H indicates the conjugate transpose operation.

[0114] In a specific embodiment, based on the principle of spatial translation invariance, the point spread function matrix is ​​generated by calculating the center point spread function vector and generating it through cyclic displacement. By utilizing the spatial symmetry of the array structure, the response characteristics of the sensor array to spatial points are fully quantified, providing a key mathematical model foundation for subsequent positioning algorithms, enabling high-resolution spatial positioning to be achieved with a small number of sensors.

[0115] The improved orthogonal matching pursuit algorithm specifically includes:

[0116] Set the initial iteration number i=0, use the original partial discharge signal as the initial residual vector, set the initial index action set to an empty set, and solve for the index of the maximum position of the inner product between the column vector of the point spread function matrix and the residual vector. The calculation expression is:

[0117] ;

[0118] Where t is the maximum position index, A H e is the conjugate transpose of the point spread function matrix. (i) Let be the residual vector after the i-th iteration;

[0119] Add the maximum position index selected in the i-th iteration to the subscript action set of the (i-1)-th iteration, calculate the generalized inverse matrix of the point spread function matrix corresponding to the subscript action set as the orthogonal projection of the point spread function matrix, solve for the new residual term vector, and calculate the new sparse approximation solution:

[0120] ;

[0121] in Let T(i) be the sparse approximation solution computed after the i-th iteration, and let A be the set of all selected scan grid point indices up to the i-th iteration. T(i) Let T(i) be the submatrix formed by the column vectors specified by T(i). For A T(i) The generalized inverse matrix, e (i-1) This is the residual vector after the (i-1)th iteration;

[0122] Let k be the preset number of iterations. When i ≤ k, the steps are repeated starting from the index of the maximum value of the inner product of the column vector and the residual vector of the solution point diffusion function matrix. When i > k, the iteration stops.

[0123] In a specific embodiment, a point spread function matrix is ​​used, and by optimizing the weight vector, the coherence between its column vectors is effectively reduced, providing a better solution dictionary for the sparse reconstruction problem, and fundamentally improving the accuracy and spatial resolution of local discharge source positioning.

[0124] In the iteration termination mechanism, an adaptive criterion based on the residual descent trend is introduced. By monitoring the rate of change of the L2 norm of the reconstructed residual, it is determined whether all local discharge sources have been captured. The iteration terminates when the contribution of adding a new atom to the signal fitting is less than a preset threshold, effectively avoiding overfitting or underfitting problems caused by an inappropriate preset number of iterations, and achieving accurate estimation of the number of discharge sources.

[0125] Step S4 specifically includes:

[0126] Step S41: Based on the sulfur hexafluoride concentration value, sulfur hexafluoride density value, coordinates of the partial discharge source and the number of discharges, the monitoring results are obtained. The monitoring results include discharge intensity, defect location and leakage severity.

[0127] The calculation of leakage severity specifically includes:

[0128] The concentration and density values ​​of sulfur hexafluoride were calculated using a piecewise linear function to obtain the concentration score and density score. The severity of the leak was then calculated by weighting the concentration score and density score.

[0129] Step S42: Determine the threshold based on the monitoring results to obtain the graded early warning results.

[0130] In a specific embodiment, the density score is obtained by comparing the sulfur hexafluoride density value with the rated value. When the density is higher than the rated value, it indicates that the insulation strength is sufficient and the risk score is 0. When the density is lower than the rated value but higher than the preset alarm value, the risk score increases linearly between 0 and 1, reflecting the degree of decrease in insulation strength. Once the density falls below the minimum safety limit, 1 point is directly assigned.

[0131] The concentration score is calculated based on the sulfur hexafluoride concentration value. When the concentration value is below the concern threshold, 0 points are awarded; when the concentration value is between the concern threshold and the danger threshold, the score increases linearly with the increase of the concentration value; when the concentration exceeds the danger threshold, 1 point is awarded.

[0132] A comprehensive leakage severity index is obtained by weighting and summing the density and concentration scores according to preset weights. The weighting follows the principle of prioritizing insulation strength, giving higher weight to the density item and lower weight to the concentration item.

[0133] Threshold judgments are applied to the monitoring results to obtain tiered early warning results, specifically including:

[0134] The tiered early warning results include the first warning, the second warning, the third warning, and the fourth warning.

[0135] When the leakage severity is less than the first threshold and no partial discharge signal is detected, no warning is triggered;

[0136] When the severity of the leak is less than the first threshold, but a partial discharge signal is detected and the discharge intensity is less than the discharge threshold, the first warning is triggered.

[0137] When the severity of the leak is less than the first threshold, but a partial discharge signal is detected and the discharge intensity is greater than the discharge threshold, a second warning is triggered.

[0138] When the severity of the leak is greater than the first threshold but less than the second threshold, and no partial discharge signal is detected, a second warning is triggered.

[0139] When the severity of the leak is greater than the first threshold but less than the second threshold, a partial discharge signal is detected, and the discharge intensity is less than the discharge threshold, a third warning is triggered.

[0140] When the severity of the leak is greater than the first threshold but less than the second threshold, a partial discharge signal is detected, and the discharge intensity is greater than the discharge threshold, the fourth warning is triggered.

[0141] If the severity of the leak exceeds the second threshold, a fourth warning will be triggered.

[0142] The first threshold is less than the second threshold.

[0143] In a specific embodiment, when the first warning is triggered, a prompt is made to perform daily monitoring of the warning location; when the second warning is triggered, a prompt is made to perform tracking and monitoring of the warning location; when the third warning is triggered, a prompt is made to strengthen the tracking and monitoring of the warning location; when the fourth warning is triggered, a prompt is made to repair the warning location; if the leakage severity is greater than the second threshold, a shutdown command is issued and a prompt is made to repair the warning location.

[0144] This invention deploys sulfur hexafluoride (SF6) sensors, density relays, and ultra-high frequency (UHF) sensors arranged in a rectangular array at key parts of equipment to simultaneously collect gas concentration, density, and partial discharge signals. For the partial discharge signals, ensemble empirical mode decomposition (EEMD) is employed. By adding Gaussian white noise and performing multiple ensemble averaging, mode aliasing is effectively suppressed, and the essential discharge mode components are stably extracted from strong noise. A transfer matrix and point spread function matrix between the scanning plane and the sensor array are established, and the orthogonal matching pursuit algorithm is sparsely reconstructed, achieving the goal of high-precision spatial positioning and quantity identification of multiple discharge sources using a small number of sensors. Finally, multi-parameter collaborative analysis and threshold judgment are performed on the gas leakage severity and the coordinates, intensity, and quantity of partial discharge sources to form a dynamic hierarchical early warning system. This system can trigger different levels of warnings based on different combinations of leakage and discharge states, significantly improving the accuracy of fault identification, the reliability of early warnings, and the intelligence level of operation and maintenance decisions.

[0145] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0146] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment, characterized in that, Includes the following steps: Step S1: Collect raw parameters; The original parameters include sulfur hexafluoride concentration value, sulfur hexafluoride density value, and original partial discharge signal; Step S2: Perform discharge preprocessing on the original partial discharge signal to obtain a partial discharge signal; Step S3: Construct the point spread function matrix and combine it with the orthogonal matching pursuit algorithm to sparsely reconstruct the partial discharge signal to obtain the coordinates of the partial discharge source and the number of discharges. Step S3 specifically includes: Step S31: Based on the propagation characteristics of partial discharge signals in sulfur hexafluoride, establish the transfer matrix between the scanning plane and the sensor array; The scanning plane is a virtual two-dimensional plane defined around the target sulfur hexafluoride insulating device. The virtual two-dimensional plane is discretized into grid points, and each grid point represents a potential discharge source location. Step S32: Construct a point spread function matrix based on the transfer matrix; Step S33: The partial discharge signal is sparsely reconstructed using an improved orthogonal matching pursuit algorithm, and the coordinates and residual vector of the partial discharge source are obtained through iteration. Step S34: Perform L2 operation on the residual vector to obtain the L2 norm of the residual vector. By monitoring the trend of the L2 norm of the residual vector with the number of iterations, the residual decrease rate is obtained. When the residual decrease rate is lower than the preset threshold, the iteration is terminated to obtain the discharge quantity of the partial discharge power source. Step S4: Based on the sulfur hexafluoride concentration value, sulfur hexafluoride density value, coordinates of the partial discharge source, and discharge quantity, a threshold judgment is made to obtain a graded early warning result.

2. The multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment as described in claim 1, characterized in that, Step S1 specifically includes: Step S11: Install a sulfur hexafluoride sensor and a density relay in the equipment components to monitor the sulfur hexafluoride concentration and density values. The equipment components include the bottom of the equipment, a flange sealing surface, a valve operating mechanism, and a density relay interface. Step S12: Set up a sensor array to receive partial discharge signals. The sensor array is a UHF sensor arranged in a rectangular pattern.

3. The multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment as described in claim 2, characterized in that, Step S2 specifically includes: Step S21: Add Gaussian white noise to the original partial discharge signal to form a noisy partial discharge signal; Step S22: Perform signal segmentation processing on the noisy partial discharge signal to obtain the IMF component and residual term; Step S23: Calculate the average value of each IMF component to obtain the IMF component. The calculation formula is as follows: ； Among them, c i (t) represents the IMF component. Let represent the i-th layer IMF component obtained from the j-th decomposition, and Ne be the overall degree parameter. The partial discharge signal is obtained by adding the IMF component to the residual term obtained from the last EMD decomposition.

4. The multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment as described in claim 3, characterized in that, The noisy partial discharge signal is subjected to signal segmentation processing, which specifically includes: The signal is divided into segments according to a preset segment length M from the starting point of the noisy partial discharge signal. The first segment is extended by m1 sampling points, and the subsequent segments are extended by m2 sampling points before and after each segment. The signal is segmented and converted into a matrix form. Each row of the matrix is ​​transformed from a segment of a noisy partial discharge signal. Each element in the matrix represents a sampled value of the noisy partial discharge signal. Each element in the matrix is ​​regarded as a key-value pair, where the key is the position of the sampling point in the noisy partial discharge signal and the value is the sampled value corresponding to the sampling point. For each segment of the noisy partial discharge signal, the extreme values ​​of the resulting matrix are determined point by point using the following method: If the value of a sampling point is greater than the values ​​of its two adjacent points, it is considered a maximum point. If the value of a sampling point is less than the values ​​of its two adjacent points, it is considered a local minimum point. The extreme points are sorted out, and the maximum and minimum value matrices are output. The upper and lower envelopes of each signal segment are calculated by cubic spline interpolation to obtain the segmented envelopes. The segmented envelopes are clipped according to the boundary positions of the preset segment length. The clipped segmented envelopes are sorted and connected by keys to output the complete envelope. The mean of the complete envelope is calculated. The noisy partial discharge signal is iteratively decomposed based on the envelope mean to extract the IMF components and residual terms of each order of the noisy partial discharge signal.

5. The multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment as described in claim 4, characterized in that, The establishment of the transfer matrix between the scanning plane and the sensor array specifically includes: The propagation delay of the partial discharge signal from the scanning grid points to each UHF sensor is calculated, and a transfer matrix is ​​constructed. The elements of the transfer matrix are represented as follows: ； Among them, g mn Let r be the element in the m-th row and n-th column of the transition matrix. mn Let ω be the distance from the m-th sensor to the n-th grid point, ω be the angular frequency, and τ be the distance from the m-th sensor to the n-th grid point. mn This corresponds to the propagation delay.

6. The multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment as described in claim 5, characterized in that, The construction of the point spread function matrix by the transition matrix specifically includes: The point spread function vector at the center of the scanning plane is calculated based on the principle of spatial translation invariance. Then, corresponding point spread function values ​​are assigned to other scanning grid points through cyclic displacement, constructing a point spread function matrix whose elements are represented as follows: ； Among them, A ij w represents the element in the i-th row and j-th column of the point spread function matrix. i Let g be the weight vector of the i-th grid point. j Let H be the array guide vector for the j-th grid point, and the superscript H indicates the conjugate transpose operation.

7. The multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment as described in claim 6, characterized in that, The improved orthogonal matching pursuit algorithm specifically includes: Set the initial iteration number i=0, use the original partial discharge signal as the initial residual vector, set the initial index action set to an empty set, and solve for the index of the maximum position of the inner product between the column vector of the point spread function matrix and the residual vector. The calculation expression is: ； Where t is the maximum position index, A H e is the conjugate transpose of the point spread function matrix. (i) Let be the residual vector after the i-th iteration; Add the maximum position index selected in the i-th iteration to the subscript action set of the (i-1)-th iteration, calculate the generalized inverse matrix of the point spread function matrix corresponding to the subscript action set as the orthogonal projection of the point spread function matrix, solve for the new residual term vector, and calculate the new sparse approximation solution: ； in Let T(i) be the sparse approximation solution computed after the i-th iteration, and let A be the set of all selected scan grid point indices up to the i-th iteration. T(i) Let T(i) be the submatrix formed by the column vectors specified by T(i). For A T(i) The generalized inverse matrix, e (i-1) This is the residual vector after the (i-1)th iteration; Let k be the preset number of iterations. When i ≤ k, the steps are repeated starting from the index of the maximum value of the inner product of the column vector and the residual vector of the solution point diffusion function matrix. When i > k, the iteration stops.

8. The multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment as described in claim 7, characterized in that, Step S4 specifically includes: Step S41: Based on the sulfur hexafluoride concentration value, sulfur hexafluoride density value, coordinates of the partial discharge source, and discharge quantity, the monitoring results are obtained, including discharge intensity, defect location, and leakage severity. The calculation of the severity of the leak specifically includes: The concentration and density values ​​of sulfur hexafluoride were calculated using a piecewise linear function to obtain the concentration score and density score. The severity of the leak was then calculated by weighting the concentration score and density score. Step S42: Determine the threshold based on the monitoring results to obtain the graded early warning results.

9. The multi-parameter collaborative monitoring method for sulfur hexafluoride insulation equipment as described in claim 8, characterized in that... The step of performing threshold judgment on the monitoring results to obtain graded early warning results specifically includes: The tiered early warning results include a first early warning, a second early warning, a third early warning, and a fourth early warning; When the leakage severity is less than the first threshold and no partial discharge signal is detected, no warning is triggered; When the severity of the leakage is less than the first threshold, but a partial discharge signal is detected and the discharge intensity is less than the discharge threshold, the first warning is triggered. When the severity of the leakage is less than the first threshold, but a partial discharge signal is detected and the discharge intensity is greater than the discharge threshold, a second warning is triggered. When the leakage severity is greater than the first threshold and less than the second threshold, and no partial discharge signal is detected, a second warning is triggered; When the leakage severity is greater than the first threshold and less than the second threshold, a partial discharge signal is detected, and the discharge intensity is less than the discharge threshold, a third warning is triggered. When the severity of the leakage is greater than the first threshold and less than the second threshold, a partial discharge signal is detected, and the discharge intensity is greater than the discharge threshold, a fourth warning is triggered. If the severity of the leak exceeds the second threshold, a fourth warning is triggered. The first threshold is less than the second threshold.

Images