Transformer bushing defect monitoring method and system based on multi-dimensional data fusion

Abstract

The invention discloses a transformer bushing defect monitoring method and system based on multi-dimensional data fusion, and the method comprises the steps: collecting a pressure data sequence and a temperature data sequence of a plurality of monitoring points in a transformer bushing, and building a multi-dimensional data matrix based on a sliding window mechanism; identifying a pressure field reconstruction event according to the multi-dimensional data matrix, and obtaining a response time difference matrix of the temperature change relative to the pressure change so as to perform time sequence correction on the temperature data sequence; establishing a dynamic correlation model among the oil level change, the pressure and the temperature by using weighted fusion, and determining a weight distribution scheme of the pressure and the temperature; and performing multi-dimensional data fusion on the pressure data sequence and the temperature data sequence which are acquired in real time according to the dynamic correlation model and the weight distribution scheme, and performing transformer bushing defect monitoring according to the obtained multi-dimensional feature vector to obtain a monitoring result. According to the invention, the accuracy of sleeve defect identification and the reliability of sleeve defect monitoring can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of transformer defect monitoring, and in particular to a transformer bushing defect monitoring method and system based on multi-dimensional data fusion. Background Art

[0002] As the core equipment of the power system, the safe and stable operation of power transformers is directly related to the reliability of the entire power system. Bushings, as key components of power transformers, undertake the important functions of high-voltage extraction and insulation protection. Real-time monitoring of their internal status is crucial for preventing equipment failures and ensuring power grid security.

[0003] Traditional casing monitoring technologies primarily rely on the independent measurement of a single parameter, lacking in-depth analysis of the coupling relationships between multidimensional parameters. For example, traditional monitoring methods often treat parameters such as pressure and temperature within the casing as independent indicators, ignoring the combined impact of oil level changes on each parameter. This results in insufficient monitoring accuracy and is prone to misjudgment. Furthermore, after obtaining multidimensional monitoring data, existing data processing algorithms often employ static fusion strategies, which are unable to adapt to the dynamic parameter variations caused by oil level fluctuations. For example, abnormal fluctuations in the casing oil level can trigger dynamic changes in the internal pressure field, a dynamic process characterized by significant spatiotemporal inhomogeneity. This dynamic pressure field further affects the heat conduction path within the casing, resulting in significant temporal differences in the responses of temperature measurement points at different locations. This ultimately leads to temporal misalignment of the multidimensional monitoring data. This data misalignment directly affects the fusion algorithm's ability to accurately identify oil level-related defect characteristics. When the algorithm cannot correctly handle this temporal inconsistency, a large number of missed detections or false alarms can occur, severely reducing the reliability of defect monitoring. Summary of the Invention

[0004] The purpose of the embodiments of the present invention is to provide a transformer bushing defect monitoring method and system based on multidimensional data fusion, which can effectively avoid misjudgment, missed detection and false alarm, achieve high-precision monitoring, and thus improve the accuracy of bushing defect identification and the reliability of bushing defect monitoring.

[0005] To achieve the above objectives, an embodiment of the present invention provides a transformer bushing defect monitoring method based on multidimensional data fusion, comprising: The pressure and temperature data sequences of multiple monitoring points inside the transformer bushing are collected according to the preset collection frequency, and a multidimensional data matrix containing the monitoring point location, collection time point, pressure and temperature is established based on the sliding window mechanism; Obtaining the pressure change rate of each monitoring point at each acquisition time point according to the multidimensional data matrix, and obtaining the starting monitoring point and starting time point of the pressure field reconstruction when determining that a pressure field reconstruction event is triggered according to the pressure change rate; Obtaining a time delay coefficient of temperature change relative to pressure change at each monitoring point based on the starting time point and the multidimensional data matrix, and establishing a response time difference matrix including monitoring point positions and time delay coefficients; Performing time series correction on the temperature data sequence in the multidimensional data matrix based on the starting monitoring point and the response time difference matrix to obtain a correction data matrix; Establishing a dynamic correlation model between oil level changes inside the transformer bushing and pressure and temperature based on the starting monitoring point and the correction data matrix, and determining a weight distribution scheme for pressure and temperature in the dynamic correlation model; According to the dynamic association model and the weight distribution scheme, multidimensional data fusion is performed on the real-time collected pressure data series and temperature data series to establish a multidimensional feature vector containing pressure, temperature and oil level changes. Transformer bushing defect monitoring is performed based on the multidimensional feature vector to obtain monitoring results; wherein the monitoring results include the defect type and its severity.

[0006] To achieve the above objectives, an embodiment of the present invention further provides a transformer bushing defect monitoring system based on multi-dimensional data fusion, comprising: A multidimensional data matrix establishment module is used to collect pressure data sequences and temperature data sequences of multiple monitoring points inside the transformer bushing according to a preset acquisition frequency, and to establish a multidimensional data matrix containing the monitoring point location, acquisition time point, pressure and temperature based on a sliding window mechanism; a pressure field reconstruction identification module, configured to obtain the pressure change rate of each monitoring point at each acquisition time point based on the multidimensional data matrix, and to obtain the starting monitoring point and starting time point of the pressure field reconstruction when a pressure field reconstruction event is determined to be triggered based on the pressure change rate; a response time difference matrix establishment module, configured to obtain a time delay coefficient of temperature change relative to pressure change at each monitoring point based on the starting time point and the multidimensional data matrix, and establish a response time difference matrix including monitoring point positions and time delay coefficients; a temperature sequence timing correction module, configured to perform timing correction on the temperature data sequence in the multidimensional data matrix based on the starting monitoring point and the response time difference matrix to obtain a correction data matrix; A correlation model and weight acquisition module is used to establish a dynamic correlation model between the oil level change inside the transformer bushing and the pressure and temperature based on the starting monitoring point and the correction data matrix, and determine the weight distribution scheme of the pressure and temperature in the dynamic correlation model; The transformer bushing defect monitoring module is used to perform multidimensional data fusion on the real-time collected pressure data sequence and temperature data sequence according to the dynamic association model and the weight distribution scheme, establish a multidimensional feature vector containing pressure, temperature and oil level changes, and perform transformer bushing defect monitoring based on the multidimensional feature vector to obtain monitoring results; wherein the monitoring results include the defect type and its severity.

[0007] Compared with the prior art, the embodiment of the present invention provides a transformer bushing defect monitoring method and system based on multidimensional data fusion. First, the pressure data sequence and temperature data sequence of multiple monitoring points inside the transformer bushing are collected according to a preset collection frequency, and a multidimensional data matrix including the monitoring point position, collection time point, pressure and temperature is established based on a sliding window mechanism; then, the pressure change rate of each monitoring point at each collection time point is obtained according to the multidimensional data matrix, and when the pressure field reconstruction event is triggered according to the pressure change rate, the starting monitoring point and starting time point of the pressure field reconstruction are obtained; then, based on the starting time point and the multidimensional data matrix, the time delay coefficient of the temperature change relative to the pressure change of each monitoring point is obtained, and a data matrix including the monitoring point position is established. and a response time difference matrix of the time delay coefficient; then, based on the starting monitoring point and the response time difference matrix, the temperature data sequence in the multidimensional data matrix is ​​time-series corrected to obtain a correction data matrix; then, based on the starting monitoring point and the correction data matrix, a dynamic correlation model between the oil level change and the pressure and temperature inside the transformer bushing is established, and the weight distribution scheme of the pressure and temperature in the dynamic correlation model is determined; finally, according to the dynamic correlation model and the weight distribution scheme, the real-time collected pressure data sequence and temperature data sequence are multidimensionally fused to establish a multidimensional feature vector containing pressure, temperature and oil level changes, and transformer bushing defect monitoring is performed based on the multidimensional feature vector to obtain monitoring results, which include the defect type and its severity. The embodiment of the present invention can effectively avoid misjudgment, missed detection and false alarm, achieve high-precision monitoring, and thus improve the accuracy of bushing defect identification and the reliability of bushing defect monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 This is a flow chart of a transformer bushing defect monitoring method based on multi-dimensional data fusion provided by one embodiment of the present invention; Figure 2 This is a structural block diagram of a transformer bushing defect monitoring system based on multi-dimensional data fusion provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0009] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this technical field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0010] The embodiment of the present invention provides a transformer bushing defect monitoring method based on multi-dimensional data fusion, see Figure 1 FIG. 1 is a flow chart of a transformer bushing defect monitoring method based on multidimensional data fusion according to an embodiment of the present invention. The method includes steps S11 to S16: Step S11: collecting pressure data sequences and temperature data sequences of multiple monitoring points inside the transformer bushing according to a preset collection frequency, and establishing a multidimensional data matrix including monitoring point positions, collection time points, pressures and temperatures based on a sliding window mechanism.

[0011] In one optional embodiment, the method of collecting pressure data sequences and temperature data sequences of multiple monitoring points inside the transformer bushing according to a preset collection frequency and establishing a multidimensional data matrix including monitoring point locations, collection time points, pressures, and temperatures based on a sliding window mechanism specifically includes: The pressure data and temperature data of multiple monitoring points inside the transformer bushing are collected according to the preset collection frequency, and the corresponding monitoring point positions and collection time points are recorded to obtain the original pressure data sequence and the original temperature data sequence; Verifying the original pressure data sequence and the original temperature data sequence to obtain a valid pressure data sequence and a valid temperature data sequence that have passed the verification; wherein the verification includes eliminating abnormal data; The valid pressure data sequence and the valid temperature data sequence are sorted and organized according to the monitoring point locations and acquisition time points, and an initial data matrix including the monitoring point locations, acquisition time points, pressure and temperature is established based on a sliding window mechanism; Data alignment processing is performed on the pressure data and temperature data of each monitoring point in the initial data matrix at the same acquisition time point to obtain a multidimensional data matrix including the monitoring point position, acquisition time point, pressure and temperature.

[0012] During specific implementation of this embodiment, the pressure data of each monitoring point (for example, pressure data can be collected using a pressure sensor) and the temperature data of each monitoring point (for example, temperature data can be collected using a temperature sensor) can be collected from multiple monitoring points inside the transformer bushing according to a preset collection frequency. The monitoring point position corresponding to each monitoring point (for example, the spatial position coordinates of the monitoring point) and the collection time point of each data collection can be recorded, and an original pressure data sequence and an original temperature data sequence containing the monitoring point position and the collection time point can be obtained accordingly. Then, the collected original pressure data sequence and original temperature data sequence can be respectively data verified to obtain a valid pressure data sequence and a valid temperature data sequence that have passed the verification. The data verification includes but is not limited to abnormal data removal. For example, the change in pressure values / temperature values ​​between adjacent collection time points in the original pressure data sequence / original temperature data sequence is calculated. If the change in pressure values / temperature values ​​exceeds a certain threshold, it is marked as abnormal data and removed from the original pressure data sequence / original temperature data sequence. Then, the valid pressure data sequence that has passed the verification is processed. The data sequence and valid temperature data sequence are sorted and organized according to the monitoring point location and acquisition time point, and an initial data matrix containing the monitoring point location, acquisition time point, pressure and temperature is established based on the sliding window mechanism. For example, the matrix only stores the data of the last 24 hours or 72 hours. When new data arrives, the new data is directly used to replace the oldest data in the matrix with the current time, that is, the old data is deleted, so as to keep the matrix size constant, and the matrix size is limited by storage capacity and computational efficiency; finally, based on the initial data matrix, the pressure data and temperature data of each monitoring point at the same acquisition time point are aligned. For example, due to the abnormal data removal processing performed in the previous article, some monitoring points may have missing pressure data / temperature data at some acquisition time points. For these missing data, the linear interpolation method can be used to calculate the value of the missing position based on the pressure value / temperature value of the corresponding monitoring point at the corresponding previous and subsequent acquisition time points, and the data is filled in the corresponding missing position. After all missing data are filled, a multidimensional data matrix containing the monitoring point location, acquisition time point, pressure and temperature can be obtained.

[0013] Among them, the construction of the multidimensional data matrix (initial data matrix) adopts a three-dimensional structure storage method. The first dimension of the multidimensional data matrix represents the location of the monitoring point, the second dimension represents the acquisition time point, and the third dimension contains two parameters: pressure value and temperature value. This matrix structure facilitates the rapid retrieval of data from a specific monitoring point at any time, and also facilitates horizontal multi-monitoring point comparative analysis and vertical time series trend analysis. Through this systematic data organization method, accurate recording and comprehensive monitoring of the operating status of the transformer bushing are achieved.

[0014] It should be noted that the arrangement of multiple monitoring points inside the transformer bushing needs to take into account the structural characteristics of the bushing and the monitoring requirements. In one possible implementation method, a monitoring point can be set every 5 meters along the axial direction of the bushing, and a pressure sensor and a temperature sensor can be installed at each monitoring point. The pressure sensor can be a piezoresistive sensor, whose working principle is to measure the pressure value by sensing the change in strain gauge resistance caused by the medium pressure inside the bushing. The temperature sensor uses a platinum resistance thermometer, whose working principle is to use the characteristic of platinum resistance changing with temperature to obtain accurate temperature readings.

[0015] It should be noted that the setting of the acquisition frequency will directly affect the timeliness and storage capacity of the data. For transformer bushings with rapidly changing pressure, the acquisition frequency can be set to once every 1 minute, while for relatively stable transformer bushings, the acquisition frequency can be reduced to once every 10 minutes. In addition, each time data is collected, the current Unix timestamp (i.e., the collection time point) can be automatically recorded with millisecond accuracy to ensure the accuracy of the data collection time stamp.

[0016] It should be noted that the data verification process is crucial to ensuring data quality. The validity of the data can be judged by calculating the data changes between adjacent time points. For example, if the pressure value collected by a monitoring point at a certain collection time point is 10 MPa, and one minute later, the pressure value collected at the next collection time point suddenly changes to 50 MPa, and the pressure change between adjacent collection time points under normal circumstances does not exceed 0.5 MPa per minute, the 50 MPa pressure data will be marked as abnormal and eliminated. A similar method is used to verify the temperature data. The normal temperature change usually does not exceed 2°C per minute. This verification mechanism can effectively prevent erroneous data caused by sensor failure or signal interference from entering the subsequent analysis process.

[0017] It should be noted that the linear interpolation method plays an important role in processing missing data. When a monitoring point lacks data at a specific collection time point, the interpolation calculation can be performed by finding the most recent valid data before and after the collection time point. For example, assuming that the pressure value of monitoring point A at 10:00 is 12 MPa, the pressure data is missing at 10:10, and the pressure value at 10:20 is 13 MPa, then the interpolation result at 10:10 is 12.5 MPa. This processing method ensures the continuity of the data and provides a complete data basis for subsequent trend analysis and defect identification.

[0018] Step S12: obtaining the pressure change rate of each monitoring point at each acquisition time point according to the multidimensional data matrix, and obtaining the starting monitoring point and starting time point of the pressure field reconstruction when determining that a pressure field reconstruction event is triggered according to the pressure change rate.

[0019] In one optional embodiment, obtaining the pressure change rate of each monitoring point at each acquisition time point according to the multidimensional data matrix, and obtaining the starting monitoring point and starting time point of the pressure field reconstruction when determining that a pressure field reconstruction event is triggered according to the pressure change rate, specifically includes: Obtaining the pressure change rate of each monitoring point at each acquisition time point according to the multidimensional data matrix; When the pressure change rate of any monitoring point at any collection time point exceeds the preset pressure fluctuation threshold, the corresponding collection time point is marked as an abnormal time point, and the corresponding monitoring point is marked as an abnormal monitoring point; When there are multiple consecutive abnormal time points at any abnormal monitoring point, a pressure field reconstruction event is determined to be triggered, and the corresponding abnormal monitoring point is used as the starting monitoring point for pressure field reconstruction, and the first abnormal time point of the starting monitoring point is used as the starting time point for pressure field reconstruction.

[0020] In combination with the above embodiments, in a specific implementation of this embodiment, a pressure data sequence of each monitoring point at each acquisition time point can be obtained from a multidimensional data matrix. For each monitoring point, the pressure change rate between adjacent acquisition time points is calculated. For example, the pressure value difference between the pressure value at the previous acquisition time point and the pressure value at the next acquisition time point can be subtracted, and the resultant pressure change rate can be divided by the time interval between the adjacent acquisition time points to obtain the pressure change rate at the corresponding acquisition time point, thereby obtaining the pressure change rate of each monitoring point at each acquisition time point. Next, the pressure change rate of each monitoring point at each acquisition time point is compared with a preset pressure fluctuation threshold value one by one. If it is determined that the pressure change rate of a monitoring point at a certain acquisition time point exceeds the preset pressure fluctuation threshold value, the acquisition time point is marked as an abnormal time point, and the monitoring point is marked as an abnormal monitoring point. Furthermore, if a certain abnormal monitoring point has multiple consecutive abnormal time points, for example, if a certain abnormal monitoring point has three consecutive abnormal time points, it is determined that the abnormal monitoring point has triggered a pressure field reconstruction event, and the abnormal monitoring point is used as the starting monitoring point for the pressure field reconstruction, and the first abnormal time point at which the starting monitoring point appears is used as the starting time point for the pressure field reconstruction.

[0021] Furthermore, the embodiment of the present invention can also determine whether there are abnormal time points at other monitoring points within a preset range of the starting monitoring point (for example, a distance range of 10 meters above and below the starting monitoring point) within a preset time window after the starting time point (for example, within 30 minutes, changes beyond this time window are not considered to be affected by the same pressure field reconstruction event). If so, it is determined that the other monitoring points are also affected by the pressure field reconstruction. At this time, the impact range of the pressure field reconstruction can be determined based on the spatial position coordinates of all other monitoring points affected by the same pressure field reconstruction event. For example, if abnormal pressure change rates are detected at monitoring points from a depth of 1000 meters to 1050 meters, the impact range is 50 meters. This quantitative impact range assessment provides accurate spatial positioning information for subsequent casing integrity evaluation and repair measure formulation, which helps to take targeted response measures in a timely manner.

[0022] It can be understood that when the pressure change rate of each monitoring point at each collection time point is compared one by one with the preset pressure fluctuation threshold, if it is determined that the pressure change rate of all monitoring points at all collection time points does not exceed the preset pressure fluctuation threshold, then it is determined that the transformer bushing is currently operating normally and there are no defects. Subsequent processing steps may not be performed, but periodic monitoring of subsequent data changes is still required.

[0023] It should be noted that when extracting pressure values ​​from a multidimensional data matrix, each monitoring point corresponds to a time series of pressure data. Assuming that the pressure value of a monitoring point is 15 MPa at 10:00 and 15.8 MPa at 10:05, and the time interval is 5 minutes, the pressure change rate is calculated as 0.8 MPa divided by 5 minutes, which is 0.16 MPa / min. This point-by-point calculation method can accurately reflect the dynamic change characteristics of pressure.

[0024] It should be noted that the determination of pressure field reconstruction events uses a continuity verification mechanism. An abnormal pressure change rate at a single acquisition time point may be caused by measurement error or transient disturbance, but the occurrence of abnormal pressure change rates at multiple consecutive acquisition time points indicates that a fundamental change in the pressure field has indeed occurred. For example, a monitoring point detects a pressure change rate of 0.35 MPa / min at 10:15, followed by pressure change rates of 0.32 MPa / min, 0.38 MPa / min, and 0.31 MPa / min at 10:20, 10:25, and 10:30, respectively. The pressure change rate at four consecutive acquisition time points exceeds the 0.3 MPa / min threshold. This continuous pressure change rate anomaly confirms the occurrence of a pressure field reconstruction event, and 10:15 can be determined as the starting time point of the pressure field reconstruction.

[0025] It should be noted that the spatial propagation characteristics of pressure field reconstruction reflect the changing laws of fluid dynamics inside the casing. When a pressure field reconstruction event occurs at a certain monitoring point, the pressure wave will propagate along the casing to the adjacent area. Assuming that monitoring point A first detects an abnormal pressure change rate at a depth of 1000 meters, its adjacent monitoring point B at a depth of 1005 meters will also experience a similar abnormal pressure change rate at a later time. By tracking this propagation process, the expansion path of the pressure field reconstruction can be accurately grasped.

[0026] Step S13: obtaining a time delay coefficient of temperature change relative to pressure change at each monitoring point based on the starting time point and the multidimensional data matrix, and establishing a response time difference matrix including monitoring point positions and time delay coefficients.

[0027] In one optional embodiment, obtaining a time delay coefficient of temperature change relative to pressure change at each monitoring point based on the starting time point and the multidimensional data matrix, and establishing a response time difference matrix including monitoring point positions and time delay coefficients, specifically includes: Extracting from the multidimensional data matrix a pressure data sequence and a temperature data sequence for each monitoring point within a preset time period before and after the starting time point; For the same monitoring point, the extracted temperature data sequence is gradually time-shifted relative to the pressure data sequence on the time axis with a preset time step. The correlation coefficient between the temperature data sequence and the pressure data sequence after each shift is calculated based on the Pearson correlation coefficient. The total shift time corresponding to the maximum value of the correlation coefficient is used as the time delay coefficient of the temperature change relative to the pressure change of the corresponding monitoring point. A response time difference matrix including the monitoring point locations and their time delay coefficients is established according to the time delay coefficients of all monitoring points.

[0028] In combination with the above embodiments, when this embodiment is specifically implemented, the pressure data sequence and temperature data sequence of each monitoring point within the preset time lengths before and after the starting time point can be extracted from the multidimensional data matrix according to the starting time point of the pressure field reconstruction; then, for the same monitoring point, the extracted temperature data sequence is gradually time-shifted relative to the extracted pressure data sequence on the time axis with a preset time step. Each time the time axis is shifted, each acquisition time point of the temperature data sequence can be superimposed with a preset time step, while the temperature value remains unchanged. For example, assuming that the temperature data sequence of a monitoring point extracted within the preset time lengths before and after the starting time point is [(t1, T1), (t2, T2) , (t3, T3)], translate once on the time axis, the preset time step is Δt, then the temperature data sequence after translation becomes [(t1+Δt, T1), (t2+Δt, T2), (t3+Δt, T3)], after each translation is completed, the correlation coefficient between the temperature data sequence after each translation and the pressure data sequence of the same monitoring point can be calculated based on the Pearson correlation coefficient, and the total translation time corresponding to the maximum value of the correlation coefficient during the entire translation process is used as the time delay coefficient of the temperature change relative to the pressure change of the same monitoring point; finally, based on the time delay coefficients of all monitoring points, a response time difference matrix containing the monitoring point positions and their corresponding time delay coefficients is established.

[0029] Among them, the rows of the response time difference matrix represent the locations of the monitoring points, and the columns represent the corresponding time delay coefficients. Assuming that there are 10 monitoring points arranged in the casing, the response time difference matrix is ​​a 10×1 column vector, where the first element corresponds to the time delay coefficient of monitoring point 1, the second element corresponds to the time delay coefficient of monitoring point 2, and so on. Through this matrix form, it is possible to quickly identify which monitoring points have abnormal response lags (this can be determined by judging whether the time delay coefficient of the monitoring point is greater than the preset time delay threshold. If so, it is determined that the corresponding monitoring point has a response lag phenomenon) and which monitoring points have response times that meet expectations. This structured data organization method not only facilitates storage and retrieval, but more importantly, provides quantitative data support for subsequent casing integrity assessment and maintenance decisions.

[0030] It should be noted that the range of data series extraction directly affects the accuracy of the analysis results. Assuming that the pressure field reconstruction event occurs at 2:30 PM, data from 2:00 PM to 3:00 PM can be extracted for analysis. For example, the pressure data series may show a sharp increase from 12 MPa to 18 MPa at 2:30 PM, while the temperature data series at monitoring points at different depths will begin to respond to this pressure change at different times. This is because when the pressure field is reconstructed, the pressure change is transmitted to different locations through the fluid medium within the casing, and temperature changes often lag behind pressure changes. Accurately identifying this lag phenomenon is crucial for understanding the thermodynamic processes within the casing. Therefore, it is necessary to perform a time shift on the temperature data series of each monitoring point on the time axis, gradually adjusting the shift amount from 0 seconds to 300 seconds. After each shift, the correlation coefficient between the temperature series and the pressure series is calculated to determine the time delay coefficient between them. The calculation process of the time delay coefficient reflects the physical properties of heat conduction.

[0031] For example, assuming that monitoring point A is closer to the pressure change source (i.e., the starting monitoring point), the correlation coefficient between its temperature data sequence and the pressure data sequence of monitoring point A reaches a maximum value of 0.92 after a translation of 15 seconds, indicating that the time delay coefficient of monitoring point A is 15 seconds; while monitoring point B, which is farther away from the pressure change source (i.e., the starting monitoring point), needs to translate its temperature data sequence for 45 seconds before the correlation coefficient between the translated temperature data sequence and the pressure data sequence of monitoring point B reaches a maximum value of 0.88, indicating that the time delay coefficient of monitoring point B is 45 seconds, indicating that it takes longer for heat to transfer to monitoring point B. This difference reflects the heat conduction path and heat transfer efficiency inside the casing.

[0032] Step S14: performing time series correction on the temperature data sequence in the multidimensional data matrix based on the starting monitoring point and the response time difference matrix to obtain a correction data matrix.

[0033] In one optional embodiment, performing time series correction on the temperature data sequence in the multidimensional data matrix based on the starting monitoring point and the response time difference matrix to obtain a corrected data matrix specifically includes: Extracting the pressure data sequence of the starting monitoring point from the multidimensional data matrix as a reference sequence, and extracting the temperature data sequence of each monitoring point as a temperature sequence to be corrected; Performing time axis translation correction on the temperature sequence to be corrected at the same monitoring point according to the time delay coefficient of each monitoring point in the response time difference matrix to obtain an intermediate temperature sequence at each monitoring point; Using the dynamic time warping algorithm, the optimal matching relationship between the intermediate temperature sequence of each monitoring point and the reference sequence at each acquisition time point is calculated respectively; According to the optimal matching relationship corresponding to each monitoring point, the intermediate temperature series of the same monitoring point is time-aligned and corrected to obtain the corrected temperature series of each monitoring point; The multidimensional data matrix is ​​updated according to the corrected temperature sequence of each monitoring point to obtain a corrected data matrix.

[0034] In combination with the above embodiments, when this embodiment is specifically implemented, the pressure data sequence of the starting monitoring point of the pressure field reconstruction can be extracted from the multidimensional data matrix as the reference sequence, and the temperature data sequence of each monitoring point can be extracted from the multidimensional data matrix as the temperature sequence to be corrected; then, a time offset is set for the temperature sequence to be corrected at the same monitoring point according to the time delay coefficient of each monitoring point in the response time difference matrix, and a time axis translation correction is performed on the corresponding temperature sequence to be corrected according to the time offset, so as to obtain the intermediate temperature sequence of each monitoring point accordingly; then, the DTW (dynamic time warping) algorithm is used to calculate the difference between the intermediate temperature sequence of each monitoring point and the reference sequence at each The optimal matching relationship of the acquisition time points, for example, can be calculated using the DTW algorithm and the Euclidean distance between the reference sequence and the intermediate temperature sequence of each monitoring point at each acquisition time point, and the dynamic programming algorithm can be gradually accumulated to obtain the minimum cumulative distance from the start point of the sequence to each position, and the optimal matching relationship between the reference sequence and the intermediate temperature sequence at each acquisition time point can be determined by backtracking the minimum cumulative distance path; then, according to the optimal matching relationship corresponding to each monitoring point, the time axis alignment correction is performed for the intermediate temperature sequence of the same monitoring point, that is, a new acquisition time point can be assigned to each temperature data in the intermediate temperature sequence according to the optimal matching relationship (for example, assuming that the optimal matching relationship shows that the temperature point t j Corresponding pressure point p i , then t j The timestamp is changed to p i timestamp), and reorganize the temperature data of each monitoring point and the pressure data at the corresponding moment according to the newly assigned collection time point, so that the temperature data and pressure data that originally had time delays are aligned on the time axis, and the corrected temperature sequence of each monitoring point is obtained accordingly; finally, the temperature data sequence of the corresponding monitoring point in the multidimensional data matrix is ​​updated according to the corrected temperature sequence of each monitoring point, and the corrected data matrix is ​​obtained accordingly.

[0035] It should be noted that the time offset is set based on the precise information provided by the response time difference matrix. Assuming that the time delay coefficient of monitoring point A is 30 seconds and the time delay coefficient of monitoring point B is 45 seconds, the temperature data series of these two monitoring points can be shifted forward by the corresponding number of seconds. This preliminary time offset lays the foundation for subsequent refined alignment. In other words, the DTW algorithm processes the temperature data series after the time delay coefficient has been initially adjusted, rather than the temperature data series extracted from the multidimensional data matrix. This "macro alignment first, then micro optimization" logic not only ensures efficiency but also improves the accuracy of timing correction, making the data from different monitoring points closer to the actual physical correspondence in the time dimension.

[0036] It should be noted that the DTW algorithm plays a key role in the time series alignment of casing monitoring data. The algorithm can handle the nonlinear time delay problem between different data series. In casing monitoring scenarios, pressure changes, as the primary physical driving factor, cause subsequent temperature changes. However, this change is not a simple fixed time delay, but rather exhibits complex time-varying characteristics depending on the fluid state and heat transfer conditions within the casing. Therefore, the DTW algorithm can be used to perform time axis correction on the time series data of each monitoring parameter and adjust the time alignment benchmark for different monitoring parameters. The calculation of Euclidean distance is the core of the DTW algorithm. For example, if the pressure value at a certain time point is 15 MPa, and the temperature series values ​​at different time points are 80°C, 82°C, and 85°C, the algorithm calculates the distance between the pressure value and each temperature value. The distance here is not a physical distance, but a measure of the degree of difference between the two parameter values. Through normalization, the two parameters of different dimensions, pressure and temperature, can be compared on the same scale.

[0037] It should be noted that the calculation process of the cumulative distance embodies the idea of ​​dynamic programming. Starting from the starting points of the two sequences, the algorithm gradually calculates the minimum cumulative distance to each position. If the current position is the i-th point in the pressure sequence and the j-th point in the temperature sequence, its cumulative distance is equal to the distance of the current point plus the minimum cumulative distance from the three possible predecessor positions to the current position. This calculation method ensures that the found path is globally optimal; the backtracking process of the optimal matching path determines the final time alignment solution. Starting from the end point of the cumulative distance matrix, the algorithm backtracks to the starting point along the direction with the fastest decrease in cumulative distance. Each point on the backtracking path represents a matching pair of the pressure sequence and the temperature sequence. For example, the 10th time point in the pressure sequence may match the 13th time point in the temperature sequence, indicating that there is a dynamic delay of 3 time units.

[0038] It should be noted that the corrected data matrix obtained after time series correction achieves true physical event alignment and can accurately reflect the causal relationship between pressure changes and temperature responses. When the pressure suddenly changes at a certain moment, the temperature data of each monitoring point can reflect the response characteristics at the corresponding moment after time correction. This alignment not only improves the accuracy of data analysis, but more importantly, provides a reliable data basis for subsequent casing status assessment and defect identification, enabling the defect identification process to more accurately identify potential problems in casing operation.

[0039] Step S15: establishing a dynamic correlation model between oil level changes inside the transformer bushing and pressure and temperature based on the starting monitoring point and the correction data matrix, and determining a weight distribution scheme for pressure and temperature in the dynamic correlation model.

[0040] In one optional embodiment, establishing a dynamic correlation model between oil level changes inside the transformer bushing and pressure and temperature based on the starting monitoring point and the correction data matrix, and determining a weight distribution scheme for pressure and temperature in the dynamic correlation model, specifically includes: According to the pressure data sequence and the corrected temperature sequence of each monitoring point in the correction data matrix, the oil level change sequence inside the transformer bushing is calculated using the principles of fluid mechanics; Calculating the correlation coefficients between the oil level change sequence, the pressure data sequence of the starting monitoring point, and the corrected temperature sequence of each monitoring point based on the Pearson correlation coefficient; Comparing each calculated correlation coefficient with a preset correlation threshold; when any correlation coefficient is less than the correlation threshold, adjusting the weight of the pressure data sequence or the corrected temperature sequence associated with the corresponding correlation coefficient to the ratio of the corresponding correlation coefficient to the correlation threshold; otherwise, keeping the initial weight unchanged, so as to obtain a weight distribution scheme for the pressure data sequence of the starting monitoring point and the corrected temperature sequence of each monitoring point; Performing weighted fusion processing on the pressure data sequence of the starting monitoring point and the corrected temperature sequence of each monitoring point according to the weight distribution scheme to obtain a comprehensive parameter sequence; The least square method is used to establish a linear correspondence between the oil level change sequence and the comprehensive parameter sequence, so as to obtain a dynamic correlation model between the oil level change inside the transformer bushing and the pressure and temperature.

[0041] In combination with the above embodiments, in a specific implementation of this embodiment, the oil level change sequence inside the transformer bushing can be inferred based on the pressure data sequence and the corrected temperature sequence of each monitoring point in the correction data matrix, using the principles of fluid mechanics (the oil level change is calculated using the combined effects of pressure change and temperature change, or it can also be directly measured by an additional oil level sensor). Then, based on the Pearson correlation coefficient, the correlation coefficient between the oil level change sequence and the pressure data sequence of the starting monitoring point for pressure field reconstruction is calculated, and the correlation coefficient between the oil level change sequence and the corrected temperature sequence of each monitoring point is calculated. For example, the correlation coefficient between the pressure data sequence of the starting monitoring point and the oil level change sequence can be obtained by calculating the covariance of the pressure data sequence and the oil level change sequence of the starting monitoring point and dividing it by the product of their standard deviations. The correlation coefficient between the corrected temperature sequence and the oil level change sequence of a certain monitoring point can be obtained by calculating the covariance of the corrected temperature sequence and the oil level change sequence of the certain monitoring point and dividing it by the product of their standard deviations. Then, each calculated correlation coefficient is compared with a preset correlation threshold. When a correlation coefficient is less than the preset correlation threshold, the pressure data sequence or the corrected temperature sequence associated with the correlation coefficient is removed. The weight of the corrected temperature sequence is adjusted to the ratio of the correlation coefficient to a preset correlation threshold. Otherwise, the weight of the pressure data sequence or the corrected temperature sequence associated with the correlation coefficient is maintained at the initial weight (the pressure data sequence of the starting monitoring point and the corrected temperature sequence of each monitoring point are both set with initial weights, for example, the initial weights are all 1), so as to obtain a weight distribution scheme for the pressure data sequence of the starting monitoring point and the corrected temperature sequence of each monitoring point, that is, to obtain the optimal fusion weight of the pressure data sequence of the starting monitoring point and the optimal fusion weight of the corrected temperature sequence of each monitoring point; then, according to the obtained weight distribution scheme, the pressure data sequence of the starting monitoring point and the corrected temperature sequence of each monitoring point are weightedly fused to obtain a comprehensive parameter sequence containing pressure and temperature; finally, the least squares method is used to establish a linear correspondence between the oil level change sequence and the comprehensive parameter sequence to obtain a dynamic correlation model between the oil level change inside the transformer bushing and the pressure and temperature. For example, the expression of the dynamic correlation model is: H = a × comprehensive parameter sequence + b, where a and b are regression coefficients, and the values ​​of the regression coefficients can be determined by minimizing the sum of squared errors between the predicted value and the actual value using the least squares method.

[0042] It's important to note that calculating the correlation coefficient is the foundation for establishing the relationship between oil level changes, pressure, and temperature. Specifically, the correlation coefficient reflects the strength of the linear relationship between two variables. In casing monitoring scenarios, oil level changes are often closely correlated with pressure changes. When pressure inside the casing increases, the fluid is compressed or pushed, causing the oil level to rise. Conversely, when pressure decreases, the oil level drops. This physical relationship can be quantified by calculating the covariance divided by the product of the standard deviations. Calculating the correlation coefficient requires synchronized time data series. Suppose, over a certain time period, the oil level rises from 2.5 meters to 3.2 meters, while the pressure increases from 10 MPa to 12 MPa. By calculating the covariance of these two series, we can find that their trends are highly consistent. A positive covariance indicates that the two variables are changing in the same direction, and a larger value indicates a stronger correlation. Dividing the covariance by the product of their respective standard deviations and normalizing the correlation coefficient gives a correlation coefficient between -1 and 1, facilitating comparisons between different parameters.

[0043] It should be noted that the setting of the correlation threshold reflects the judgment criteria for the importance of parameters. There are differences in the degree of correlation between different parameters and oil level. Pressure parameters usually have a direct physical causal relationship with oil level changes, and the correlation coefficient often reaches above 0.8. The influence of temperature parameters is relatively indirect, mainly affecting the oil level by changing the fluid viscosity and density, and the correlation coefficient may be around 0.5. Preferably, the correlation threshold can be set to 0.6. Parameters associated with correlation coefficients above this correlation threshold are considered to have significant influence, and parameters associated with correlation coefficients below this correlation threshold need to reduce their weights in the dynamic correlation model. The weight adjustment mechanism ensures the accuracy and stability of the dynamic correlation model.

[0044] For example, assuming that the correlation coefficient corresponding to the corrected temperature sequence of a monitoring point is 0.45, which is lower than the correlation threshold of 0.6, its weight will be adjusted to 0.45 divided by 0.6, that is, 0.75, which means that the corrected temperature sequence of the monitoring point only contributes 75% of the original value in the final comprehensive parameter calculation; assuming that the correlation coefficient corresponding to the pressure data sequence of the starting monitoring point is 0.85, which is higher than the correlation threshold of 0.6, its weight will be maintained at 1, fully reflecting its dominant role in the oil level change.

[0045] It should be noted that the implementation of the weighted fusion algorithm reflects the concept of the combined influence of multiple parameters. For example, assume that at a certain moment, the pressure value is 11 MPa, with a weight of 1; the temperature value 1 is 82°C, with a weight of 0.75; and the temperature value 2 is 85°C, with a weight of 0.9. The weighted fusion parameter value is calculated as: 11 × 1 + 82 × 0.75 + 85 × 0.9. The resulting composite parameter value reflects the weighted influence of multiple parameters. This approach preserves the role of the main influencing factors while appropriately considering the contribution of secondary factors.

[0046] It should be noted that the establishment of the dynamic correlation model adopts the linear relationship assumption, and the optimal regression coefficient can be determined by minimizing the deviation between the actual oil level value and the model predicted value. The advantages of this method are simple model form, high computational efficiency, and clear physical meaning. The model established in this way can accurately predict the oil level changes under given pressure and temperature conditions, providing reliable technical support for real-time monitoring of the casing operation status and abnormal warning.

[0047] Step S16: Based on the dynamic association model and the weight distribution scheme, multidimensional data fusion is performed on the real-time collected pressure data sequence and temperature data sequence to establish a multidimensional feature vector containing pressure, temperature and oil level changes, and transformer bushing defect monitoring is performed based on the multidimensional feature vector to obtain monitoring results; wherein the monitoring results include the defect type and its severity.

[0048] In one optional embodiment, the method of performing multidimensional data fusion on the real-time collected pressure data sequence and temperature data sequence according to the dynamic association model and the weight distribution scheme, establishing a multidimensional feature vector including pressure, temperature and oil level changes, and performing transformer bushing defect monitoring based on the multidimensional feature vector to obtain monitoring results specifically includes: performing weighted processing on the pressure data sequence of the starting monitoring point and the corrected temperature sequence of each monitoring point according to the weight distribution scheme to obtain a weighted pressure feature of the starting monitoring point and a weighted temperature feature of each monitoring point; Inputting the weighted pressure characteristic of the starting monitoring point and the weighted temperature characteristic of each monitoring point into the dynamic correlation model to obtain the oil level change characteristic; Establishing a standard feature vector based on the weighted pressure feature of the starting monitoring point, the weighted temperature feature of each monitoring point, and the oil level change feature; Based on the structure of the standard feature vector, multi-dimensional data fusion is performed on the real-time collected pressure data sequence and temperature data sequence to establish a multi-dimensional feature vector containing pressure, temperature and oil level changes; Inputting the multidimensional feature vector into a trained defect recognition model to obtain the confidence level of each defect category identified by the multidimensional feature vector; Based on the confidence level of each defect category identified by the multidimensional feature vector, it is determined whether the transformer bushing has a defect type corresponding to the defect category. If so, the severity of the defect type is determined based on the corresponding confidence level, so as to obtain a monitoring result based on the existing defect type and its severity.

[0049] In combination with the above embodiment, when this embodiment is specifically implemented, the pressure data sequence of the starting monitoring point and the corrected temperature sequence of each monitoring point can be weighted according to the obtained weight distribution scheme (including the optimal fusion weight of the pressure data sequence of the starting monitoring point and the optimal fusion weight of the corrected temperature sequence of each monitoring point), and the weighted pressure feature of the starting monitoring point (obtained by multiplying the pressure data sequence of the starting monitoring point by its optimal fusion weight) and the weighted temperature feature of each monitoring point (obtained by multiplying the corrected temperature sequence of each monitoring point by the corresponding optimal fusion weight) can be obtained accordingly; then, the obtained starting monitoring point The weighted pressure feature of the starting monitoring point and the weighted temperature feature of each monitoring point are input into the dynamic correlation model to obtain the corresponding oil level change feature; then, a standard feature vector is established based on the weighted pressure feature of the starting monitoring point, the weighted temperature feature of each monitoring point and the oil level change feature. For example, the standard feature vector can be expressed as a vector form of [weighted pressure value, weighted temperature value 1, weighted temperature value 2, ..., oil level change value]; then, the feature structure of the standard feature vector is used as a standard to perform multi-dimensional data fusion on the real-time collected pressure data sequence and temperature data sequence (that is, the real-time collected pressure data sequence and temperature data sequence are fused into a multi-dimensional data structure). The processing corresponding to steps S11 to S14 is executed sequentially, and then the multidimensional data fusion in this embodiment is performed based on the processing results) to establish a multidimensional feature vector that is consistent with the feature structure of the standard feature vector and includes real-time pressure, real-time temperature, and real-time oil level changes; then, the obtained multidimensional feature vector is input into the trained defect recognition model for recognition, and the confidence level of each defect category identified by the multidimensional feature vector is obtained accordingly; finally, based on the confidence level of each defect category identified by the multidimensional feature vector, it is determined whether the transformer bushing has the defect type corresponding to the defect category, for example, to determine the position of a defect category. Whether the confidence is higher than the preset identification threshold, if so, it is determined that the transformer bushing has the defect type corresponding to the defect category; otherwise, it is determined that the transformer bushing does not have the defect type corresponding to the defect category. Accordingly, if it is determined that the transformer bushing has the defect type corresponding to the defect category, the severity of the defect type corresponding to the defect category is determined according to the numerical range of the confidence of the defect category, so as to obtain the monitoring result according to the existing defect type (for example, normal state, leakage defect, valve failure, pressure abnormality, etc.) and its corresponding severity (for example, mild leakage, moderate leakage, severe leakage, etc.).

[0050] It should be noted that the construction of multidimensional feature vectors embodies the core idea of ​​multidimensional data fusion. According to the weight distribution scheme determined in the previous article, if the optimal fusion weight of the temperature value is 1 and the optimal fusion weight of the temperature value is 0.8, then when constructing the multidimensional feature vector, the pressure value directly uses the original value, and the temperature value needs to be multiplied by a weight coefficient of 0.8. This weighted processing ensures that the parameters that have a greater impact on oil level changes occupy a more important position in the feature space, thereby improving the accuracy of subsequent defect category classification.

[0051] It should be noted that for the trained defect recognition model, feature vectors under normal operating conditions and feature vectors when oil level abnormalities occur (consistent with the feature structure of standard feature vectors) can be extracted from historical monitoring data, and corresponding category labels can be marked for each feature vector to form a training data set; the training data set is input into the support vector machine classification algorithm, and the feature vectors are mapped to a high-dimensional space through the kernel function. The decision boundary that can distinguish different categories is found in the high-dimensional space, and the defect recognition model is trained. The model calculates the confidence value of the feature vector belonging to each category based on the distance between the feature vector and the decision boundary. The multidimensional feature vector collected in real time and weighted is input into the trained defect recognition model, and the model will calculate the distance from the multidimensional feature vector to the decision boundary of each category and convert it into a confidence value.

[0052] It should be noted that the construction of the training data set requires sufficient historical data support. Under normal operating conditions, the oil level fluctuates stably within the range of 2.5 meters to 3.5 meters, the corresponding pressure value is between 10MPa and 12MPa, and the temperature is between 80℃ and 85℃. When an abnormal oil level occurs, such as a sudden drop in the oil level to below 2 meters, it is usually accompanied by an abnormal pressure drop to below 8MPa. These abnormal state data are marked as corresponding defect types, such as "leakage defect" or "valve failure". The kernel function mapping of the support vector machine is the key to achieving nonlinear classification. In the original feature space, the data points of the normal state and the defect state may overlap, which is difficult to separate with a simple linear boundary. The path of the support vector machine is used to map the normal state and the defect state. The basis kernel function maps the three-dimensional original feature vector to a higher-dimensional space, making the originally overlapping data points separable in the new space. This mapping maintains the relative relationship between data points while enhancing the distinction between different categories. The process of determining the decision boundary reflects the optimization idea of ​​the support vector machine. The algorithm searches for the hyperplane that can maximize the interval between different categories. The data points closest to the decision boundary are called support vectors, which determine the position of the boundary. For example, the support vectors of the normal state may be data points with a pressure of 11.5MPa and a temperature of 83℃, while the support vectors of the leakage defect may be data points with a pressure of 7.5MPa and a temperature of 78℃. These key data points determine the precise position of the classification boundary.

[0053] It should be noted that the calculation of confidence is based on the distance from the feature vector to the decision boundary. Preferably, when a new feature vector is input into the model, the algorithm calculates the algebraic distance from the point to the decision boundary of each category. A positive distance indicates that it is on the positive side of the boundary, and a negative distance indicates that it is on the negative side. These distance values ​​can then be converted into confidence in the form of probabilities between 0 and 1 through the sigmoid function. For example, a confidence of 0.9 means that the model is 90% sure that the sample belongs to a certain defect category.

[0054] It should be noted that the severity can be determined using a segmented mapping method. For example, for leakage defects, a confidence level between 0.7 and 0.8 is considered a mild leakage, between 0.8 and 0.9 is considered a moderate leakage, and above 0.9 is considered a severe leakage. This grading method enables maintenance personnel to formulate corresponding treatment strategies based on the severity of the defect. Mild defects can be repaired in a planned manner, while severe defects require immediate treatment, thereby achieving refined management of casing operation risks.

[0055] In one optional embodiment, the method further includes: A defect evolution trend analysis is performed on the defect types included in the monitoring results according to a historical monitoring database, and a corresponding defect early warning mechanism is triggered according to the analysis results and the monitoring results.

[0056] In one optional embodiment, performing defect evolution trend analysis on the defect types included in the monitoring results based on the historical monitoring database, and triggering a corresponding defect early warning mechanism based on the analysis results and the monitoring results, specifically includes: Taking the defect type included in the monitoring result as the target defect, extracting historical record data corresponding to defects of the same type as the target defect from the historical monitoring database; wherein the historical record data includes each time point when the defect of the same type was identified and its severity; Perform defect evolution trend analysis based on the historical record data to obtain the average evolution speed of the target defect; Taking the severity of the target defect included in the monitoring result as the current severity, and predicting the deterioration time required for the target defect to reach the maximum severity from the current severity according to the average evolution rate; A corresponding defect warning mechanism is triggered according to the deterioration time and the current severity; wherein the defect warning mechanism is divided into different warning levels based on different deterioration times and different current severity.

[0057] In combination with the above embodiments, when this embodiment is specifically implemented, the defect type included in the monitoring results can be used as the target defect, and the historical record data corresponding to the defect of the same defect type as the target defect (i.e., the same type of defect) can be extracted from the historical monitoring database. The historical record data includes each time point when the same type of defect is identified and its severity, especially the time point and severity when the same type of defect is first discovered, as well as the evolution state at subsequent time points; then, based on the extracted historical record data, the defect evolution trend analysis is performed on the same type of defect, and the average evolution speed of the target defect is obtained accordingly. For example, the same type of defects often have similar evolution laws. Based on the extracted historical record data, a linear regression method can be used to take time as the independent variable to analyze the evolution trend of the same type of defects. The severity value of the defect is fitted as the dependent variable, and the relationship between the severity value and time is obtained accordingly. The slope of the fitted straight line can be used as the average evolution speed of defects of the same type, which is equivalent to the average evolution speed of the target defect. Then, the severity of the target defect contained in the monitoring result is used as the current severity. According to the average evolution speed and current severity of the target defect, the deterioration time required for the target defect to reach the highest severity from the current severity is predicted, and the defect deterioration time prediction value is obtained accordingly. Finally, according to the obtained defect deterioration time prediction value and the current severity of the target defect, the corresponding defect warning mechanism is triggered, wherein the defect warning mechanism is divided into different warning levels based on different deterioration times and different current severities.

[0058] For example, the relationship between severity and time, obtained using linear regression, can be expressed as: y = kx + c, where x represents time, y represents severity, k represents slope, and c represents intercept. The slope k and intercept c can be obtained using the least squares method. For example, if the fitted slope for a leak defect is 8 units per month, this means that the severity increases by 8 units per month on average.

[0059] It should be noted that the severity value can be quantitatively expressed in percentage. For example, a severity value in the range of 0 to 30 indicates a mild defect, a severity value in the range of 30 to 70 indicates a moderate defect, and a severity value in the range of 70 to 100 indicates a severe defect. Among them, a severity value of 100 represents the highest severity, that is, the defect has developed to a critical state that must be dealt with immediately. This quantification method makes the severity of different types of defects comparable, which facilitates unified management and analysis.

[0060] It should be noted that the prediction of deterioration time is calculated based on the current state and evolution rate. For example, assuming that the current severity of a defect is 45, which is a moderate defect, and the average evolution rate obtained based on historical monitoring data is 8 units per month, then the difference from 45 to 100 is 55, divided by the average evolution rate of 8, resulting in a deterioration time of approximately 7 months. This predicted value indicates that if no intervention measures are taken, the defect will develop into the most serious state in 7 months.

[0061] It should be noted that the design of the warning level of the defect warning mechanism reflects the refined requirements of risk management. For example, the warning levels of the complete defect warning mechanism are divided into: Level 1 warning: deterioration time < 3 months and current severity is moderate (30-70) or severe (70-100); Level 2 warning: The deterioration time is between 3 and 6 months and the current severity is moderate (30-70) or severe (70-100); Level 3 warning: The deterioration time is between 6 and 12 months and the current severity is moderate (30-70) or severe (70-100); Level 4 warning: deterioration time > 12 months and current severity is moderate (30-70); Monitoring only (which can be understood as a level zero warning): The current severity is mild (0-30), and regular checks are performed but no warning is triggered.

[0062] For example, when the deterioration time is less than 3 months and the defect is currently a moderate defect, it means that the defect is deteriorating rapidly and the basic severity is high, and a first-level warning needs to be triggered, requiring immediate maintenance. When the deterioration time is between 3 and 6 months and the defect is currently a moderate defect, a second-level warning needs to be triggered, and it can be included in the planned maintenance arrangement.

[0063] It should be noted that this defect evolution trend analysis and early warning mechanism based on historical monitoring data has important practical value. By accurately predicting the deterioration time of defects, maintenance personnel can reasonably arrange maintenance plans and intervene before the defects develop to a dangerous level. At the same time, graded early warning also avoids excessive investment of resources. For defects that evolve slowly, regular monitoring can be adopted, while defects that deteriorate rapidly can be responded to in a timely manner, realizing an organic combination of preventive maintenance and emergency response.

[0064] A transformer bushing defect monitoring method based on multidimensional data fusion provided by an embodiment of the present invention collects pressure data sequences and temperature data sequences from multiple monitoring points inside the transformer bushing, establishes a multidimensional data matrix based on a sliding window mechanism, identifies pressure field reconstruction events based on the multidimensional data matrix, and obtains a response time difference matrix of temperature changes relative to pressure changes to perform time series correction on the temperature data sequence, then uses weighted fusion to establish a dynamic correlation model between oil level changes and pressure and temperature, and determines a weight distribution scheme for pressure and temperature, and then performs multidimensional data fusion on the real-time collected pressure data sequences and temperature data sequences based on the dynamic correlation model and the weight distribution scheme. , and monitor transformer bushing defects based on the obtained multi-dimensional feature vector to obtain monitoring results. It also analyzes the defect evolution trend to obtain the average evolution speed of the defect, and predicts the deterioration time required for the defect to reach the highest severity from the current severity based on the average evolution speed of the defect. Then, according to the deterioration time and the current severity of the defect, different levels of defect early warning mechanisms are triggered, thereby effectively avoiding misjudgment, missed detection and false alarms, achieving high-precision monitoring, and thus improving the accuracy of casing defect identification and the reliability of casing defect monitoring. At the same time, it can also accurately monitor the oil level changes in the casing, timely identify potential defects, effectively prevent safety accidents, and improve the reliability of oil well operation.

[0065] The embodiment of the present invention further provides a transformer bushing defect monitoring system based on multi-dimensional data fusion, which is used to implement the transformer bushing defect monitoring method based on multi-dimensional data fusion described in any of the above embodiments. Figure 2 FIG. 1 is a block diagram of a transformer bushing defect monitoring system based on multidimensional data fusion according to an embodiment of the present invention. The system includes: A multidimensional data matrix establishment module 11 is used to collect pressure data sequences and temperature data sequences of multiple monitoring points inside the transformer bushing according to a preset acquisition frequency, and establish a multidimensional data matrix including the monitoring point location, acquisition time point, pressure and temperature based on a sliding window mechanism; a pressure field reconstruction identification module 12 for obtaining the pressure change rate of each monitoring point at each acquisition time point based on the multidimensional data matrix, and obtaining the starting monitoring point and starting time point of the pressure field reconstruction when a pressure field reconstruction event is determined to be triggered based on the pressure change rate; a response time difference matrix establishing module 13, configured to obtain a time delay coefficient of temperature change relative to pressure change at each monitoring point based on the starting time point and the multidimensional data matrix, and establish a response time difference matrix including monitoring point positions and time delay coefficients; a temperature sequence timing correction module 14 for performing timing correction on the temperature data sequence in the multidimensional data matrix based on the starting monitoring point and the response time difference matrix to obtain a correction data matrix; A correlation model and weight acquisition module 15 is used to establish a dynamic correlation model between the oil level change inside the transformer bushing and the pressure and temperature based on the starting monitoring point and the correction data matrix, and determine a weight distribution scheme for the pressure and temperature in the dynamic correlation model; The transformer bushing defect monitoring module 16 is configured to perform multidimensional data fusion on the real-time collected pressure data sequence and temperature data sequence according to the dynamic association model and the weight distribution scheme, establish a multidimensional feature vector containing pressure, temperature and oil level changes, and perform transformer bushing defect monitoring based on the multidimensional feature vector to obtain monitoring results; wherein the monitoring results include the defect type and its severity.

[0066] Preferably, the multidimensional data matrix establishing module 11 specifically includes: The raw data sequence acquisition unit is used to collect pressure data and temperature data of multiple monitoring points inside the transformer bushing according to a preset acquisition frequency, and record the corresponding monitoring point positions and acquisition time points to obtain raw pressure data sequences and raw temperature data sequences; an original data sequence verification unit, configured to verify the original pressure data sequence and the original temperature data sequence to obtain a valid pressure data sequence and a valid temperature data sequence that have passed verification; wherein the verification includes eliminating abnormal data; an initial data matrix establishing unit, configured to sort and organize the valid pressure data sequence and the valid temperature data sequence according to the monitoring point positions and the acquisition time points, and establish an initial data matrix including the monitoring point positions, the acquisition time points, the pressure and the temperature based on a sliding window mechanism; The multidimensional data matrix establishment unit is used to perform data alignment processing on the pressure data and temperature data of each monitoring point in the initial data matrix at the same acquisition time point to obtain a multidimensional data matrix including the monitoring point position, acquisition time point, pressure and temperature.

[0067] Preferably, the pressure field reconstruction and identification module 12 specifically includes: a pressure change rate acquisition unit, configured to acquire the pressure change rate of each monitoring point at each acquisition time point according to the multidimensional data matrix; An abnormal time point and monitoring point identification unit is used to mark the corresponding collection time point as an abnormal time point and the corresponding monitoring point as an abnormal monitoring point when the pressure change rate of any monitoring point at any collection time point exceeds a preset pressure fluctuation threshold; The pressure field reconstruction identification unit is used to determine the triggering of a pressure field reconstruction event when there are multiple consecutive abnormal time points at any abnormal monitoring point, and to use the corresponding abnormal monitoring point as the starting monitoring point for pressure field reconstruction, and the first abnormal time point of the starting monitoring point as the starting time point for pressure field reconstruction.

[0068] Preferably, the response time difference matrix establishing module 13 specifically includes: a first data sequence extraction unit, configured to extract, from the multidimensional data matrix, a pressure data sequence and a temperature data sequence for each monitoring point within a preset time period before and after the starting time point; a time delay coefficient acquisition unit, for performing a stepwise time shift of the extracted temperature data sequence relative to the pressure data sequence on the time axis at a preset time step for the same monitoring point, and calculating the correlation coefficient between the temperature data sequence and the pressure data sequence after each shift based on the Pearson correlation coefficient, and taking the total shift time corresponding to the maximum value of the correlation coefficient as the time delay coefficient of the temperature change relative to the pressure change of the corresponding monitoring point; The response time difference matrix establishing unit is used to establish a response time difference matrix including the monitoring point positions and their time delay coefficients according to the time delay coefficients of all monitoring points.

[0069] Preferably, the temperature sequence timing correction module 14 specifically includes: a second data sequence extraction unit, configured to extract the pressure data sequence of the starting monitoring point from the multidimensional data matrix as a reference sequence, and extract the temperature data sequence of each monitoring point as a temperature sequence to be corrected; A first timing correction unit is configured to perform time axis translation correction on the temperature sequence to be corrected at the same monitoring point according to the time delay coefficient of each monitoring point in the response time difference matrix, so as to obtain an intermediate temperature sequence at each monitoring point; An optimal matching relationship acquisition unit, configured to calculate the optimal matching relationship between the intermediate temperature sequence of each monitoring point and the reference sequence at each acquisition time point using a dynamic time warping algorithm; The second time series correction unit is used to perform time axis alignment correction on the intermediate temperature sequence of the same monitoring point according to the optimal matching relationship corresponding to each monitoring point, so as to obtain a corrected temperature sequence of each monitoring point; The multidimensional data matrix updating unit is used to update the multidimensional data matrix according to the corrected temperature sequence of each monitoring point to obtain a corrected data matrix.

[0070] Preferably, the association model and weight acquisition module 15 specifically includes: An oil level change sequence estimation unit, configured to estimate the oil level change sequence inside the transformer bushing using the principles of fluid mechanics based on the pressure data sequence and the corrected temperature sequence of each monitoring point in the correction data matrix; a correlation coefficient calculation unit, configured to calculate the correlation coefficients between the oil level change sequence and the pressure data sequence of the starting monitoring point and the corrected temperature sequence of each monitoring point based on the Pearson correlation coefficient; a weight distribution scheme acquisition unit, configured to compare each calculated correlation coefficient with a preset correlation threshold value; when any correlation coefficient is less than the correlation threshold value, adjust the weight of the pressure data sequence or the corrected temperature sequence associated with the corresponding correlation coefficient to the ratio of the corresponding correlation coefficient to the correlation threshold value; otherwise, keep the initial weight unchanged, so as to obtain a weight distribution scheme for the pressure data sequence of the starting monitoring point and the corrected temperature sequence of each monitoring point; a comprehensive parameter sequence acquisition unit, configured to perform weighted fusion processing on the pressure data sequence of the starting monitoring point and the corrected temperature sequence of each monitoring point according to the weight distribution scheme to obtain a comprehensive parameter sequence; The dynamic correlation model acquisition unit is used to establish a linear correspondence between the oil level change sequence and the comprehensive parameter sequence using the least square method to obtain a dynamic correlation model between the oil level change and the pressure and temperature inside the transformer bushing.

[0071] Preferably, the transformer bushing defect monitoring module 16 specifically includes: a pressure and temperature feature acquisition unit, configured to perform weighted processing on the pressure data sequence of the starting monitoring point and the corrected temperature sequence of each monitoring point according to the weight distribution scheme, to obtain a weighted pressure feature of the starting monitoring point and a weighted temperature feature of each monitoring point; an oil level variation characteristic acquisition unit, configured to input the weighted pressure characteristic of the starting monitoring point and the weighted temperature characteristic of each monitoring point into the dynamic correlation model to obtain the oil level variation characteristic; a standard feature vector establishing unit, configured to establish a standard feature vector based on the weighted pressure feature of the starting monitoring point, the weighted temperature feature of each monitoring point, and the oil level change feature; A multidimensional feature vector establishment unit is used to perform multidimensional data fusion on the real-time collected pressure data sequence and temperature data sequence based on the structure of the standard feature vector, and establish a multidimensional feature vector containing pressure, temperature and oil level changes; a defect category identification unit, configured to input the multidimensional feature vector into a trained defect identification model to obtain a confidence level that the multidimensional feature vector is identified as each defect category; The defect monitoring result acquisition unit is used to determine whether the transformer bushing has a defect type corresponding to the defect category based on the confidence level of each defect category identified by the multidimensional feature vector; if so, determine the severity of the defect type based on the corresponding confidence level, so as to obtain a monitoring result based on the existing defect type and its severity.

[0072] Preferably, the system further comprises: The defect evolution classification and early warning module is used to analyze the defect evolution trend of the defect types contained in the monitoring results based on the historical monitoring database, and trigger the corresponding defect early warning mechanism based on the analysis results and the monitoring results.

[0073] Preferably, the defect evolution classification and early warning module specifically includes: a historical record data extraction unit, configured to extract historical record data corresponding to defects of the same type as the target defect, using the defect type included in the monitoring result as the target defect, from a historical monitoring database; wherein the historical record data includes each time point at which the defect of the same type was identified and its severity; A defect evolution rate acquisition unit, configured to perform defect evolution trend analysis based on the historical record data to obtain an average evolution rate of the target defect; a defect deterioration time prediction unit, configured to use the severity of the target defect included in the monitoring result as the current severity and predict the deterioration time required for the target defect to reach a maximum severity from the current severity according to the average evolution rate; The defect warning triggering unit is used to trigger the corresponding defect warning mechanism according to the deterioration time and the current severity; wherein the defect warning mechanism is divided into different warning levels based on different deterioration times and different current severity.

[0074] It should be noted that the transformer bushing defect monitoring system based on multidimensional data fusion provided in an embodiment of the present invention can implement all the processes of the transformer bushing defect monitoring method based on multidimensional data fusion described in any of the above embodiments. The functions and technical effects achieved by each module and unit in the system are respectively the same as the functions and technical effects achieved by the transformer bushing defect monitoring method based on multidimensional data fusion described in the above embodiments, and will not be repeated here.

[0075] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A transformer bushing defect monitoring method based on multidimensional data fusion, characterized in that: include: The pressure and temperature data sequences of multiple monitoring points inside the transformer bushing are collected according to the preset collection frequency, and a multidimensional data matrix containing the monitoring point location, collection time point, pressure and temperature is established based on the sliding window mechanism; Obtaining the pressure change rate of each monitoring point at each acquisition time point according to the multidimensional data matrix, and obtaining the starting monitoring point and starting time point of the pressure field reconstruction when determining that a pressure field reconstruction event is triggered according to the pressure change rate; Obtaining a time delay coefficient of temperature change relative to pressure change at each monitoring point based on the starting time point and the multidimensional data matrix, and establishing a response time difference matrix including monitoring point positions and time delay coefficients; Performing time series correction on the temperature data sequence in the multidimensional data matrix based on the starting monitoring point and the response time difference matrix to obtain a correction data matrix; Establishing a dynamic correlation model between oil level changes inside the transformer bushing and pressure and temperature based on the starting monitoring point and the correction data matrix, and determining a weight distribution scheme for pressure and temperature in the dynamic correlation model; According to the dynamic association model and the weight distribution scheme, multidimensional data fusion is performed on the real-time collected pressure data series and temperature data series to establish a multidimensional feature vector containing pressure, temperature and oil level changes. Transformer bushing defect monitoring is performed based on the multidimensional feature vector to obtain monitoring results; wherein the monitoring results include the defect type and its severity.

2. The transformer bushing defect monitoring method based on multidimensional data fusion according to claim 1, characterized in that: The method collects pressure data sequences and temperature data sequences of multiple monitoring points inside the transformer bushing according to a preset collection frequency, and establishes a multidimensional data matrix including monitoring point locations, collection time points, pressures, and temperatures based on a sliding window mechanism, specifically including: The pressure data and temperature data of multiple monitoring points inside the transformer bushing are collected according to the preset collection frequency, and the corresponding monitoring point positions and collection time points are recorded to obtain the original pressure data sequence and the original temperature data sequence; Verifying the original pressure data sequence and the original temperature data sequence to obtain a valid pressure data sequence and a valid temperature data sequence that have passed the verification; wherein the verification includes eliminating abnormal data; The valid pressure data sequence and the valid temperature data sequence are sorted and organized according to the monitoring point locations and acquisition time points, and an initial data matrix including the monitoring point locations, acquisition time points, pressure and temperature is established based on a sliding window mechanism; Data alignment processing is performed on the pressure data and temperature data of each monitoring point in the initial data matrix at the same acquisition time point to obtain a multidimensional data matrix including the monitoring point position, acquisition time point, pressure and temperature.

3. The transformer bushing defect monitoring method based on multidimensional data fusion according to claim 1, characterized in that: The step of obtaining the pressure change rate of each monitoring point at each acquisition time point according to the multidimensional data matrix, and obtaining the starting monitoring point and starting time point of the pressure field reconstruction when determining that a pressure field reconstruction event is triggered according to the pressure change rate, specifically includes: Obtaining the pressure change rate of each monitoring point at each acquisition time point according to the multidimensional data matrix; When the pressure change rate of any monitoring point at any collection time point exceeds the preset pressure fluctuation threshold, the corresponding collection time point is marked as an abnormal time point, and the corresponding monitoring point is marked as an abnormal monitoring point; When there are multiple consecutive abnormal time points at any abnormal monitoring point, a pressure field reconstruction event is determined to be triggered, and the corresponding abnormal monitoring point is used as the starting monitoring point for pressure field reconstruction, and the first abnormal time point of the starting monitoring point is used as the starting time point for pressure field reconstruction.

4. The transformer bushing defect monitoring method based on multidimensional data fusion according to claim 1, characterized in that: The step of obtaining a time delay coefficient of temperature change relative to pressure change at each monitoring point based on the starting time point and the multidimensional data matrix, and establishing a response time difference matrix including monitoring point positions and time delay coefficients, specifically includes: Extracting from the multidimensional data matrix a pressure data sequence and a temperature data sequence for each monitoring point within a preset time period before and after the starting time point; For the same monitoring point, the extracted temperature data sequence is gradually time-shifted relative to the pressure data sequence on the time axis with a preset time step. The correlation coefficient between the temperature data sequence and the pressure data sequence after each shift is calculated based on the Pearson correlation coefficient. The total shift time corresponding to the maximum value of the correlation coefficient is used as the time delay coefficient of the temperature change relative to the pressure change of the corresponding monitoring point. A response time difference matrix including the monitoring point locations and their time delay coefficients is established according to the time delay coefficients of all monitoring points.

5. The transformer bushing defect monitoring method based on multidimensional data fusion according to claim 1, characterized in that: The performing time series correction on the temperature data sequence in the multidimensional data matrix based on the starting monitoring point and the response time difference matrix to obtain a correction data matrix specifically includes: Extracting the pressure data sequence of the starting monitoring point from the multidimensional data matrix as a reference sequence, and extracting the temperature data sequence of each monitoring point as a temperature sequence to be corrected; Performing time axis translation correction on the temperature sequence to be corrected at the same monitoring point according to the time delay coefficient of each monitoring point in the response time difference matrix to obtain an intermediate temperature sequence at each monitoring point; Using the dynamic time warping algorithm, the optimal matching relationship between the intermediate temperature sequence of each monitoring point and the reference sequence at each acquisition time point is calculated respectively; According to the optimal matching relationship corresponding to each monitoring point, the intermediate temperature series of the same monitoring point is time-aligned and corrected to obtain the corrected temperature series of each monitoring point; The multidimensional data matrix is ​​updated according to the corrected temperature sequence of each monitoring point to obtain a corrected data matrix.

6. The transformer bushing defect monitoring method based on multidimensional data fusion according to claim 5, characterized in that: The step of establishing a dynamic correlation model between the oil level change inside the transformer bushing and the pressure and temperature based on the starting monitoring point and the correction data matrix, and determining a weight distribution scheme for the pressure and temperature in the dynamic correlation model, specifically includes: According to the pressure data sequence and the corrected temperature sequence of each monitoring point in the correction data matrix, the oil level change sequence inside the transformer bushing is calculated using the principles of fluid mechanics; Calculating the correlation coefficients between the oil level change sequence, the pressure data sequence of the starting monitoring point, and the corrected temperature sequence of each monitoring point based on the Pearson correlation coefficient; Comparing each calculated correlation coefficient with a preset correlation threshold; when any correlation coefficient is less than the correlation threshold, adjusting the weight of the pressure data sequence or the corrected temperature sequence associated with the corresponding correlation coefficient to the ratio of the corresponding correlation coefficient to the correlation threshold; otherwise, keeping the initial weight unchanged, so as to obtain a weight distribution scheme for the pressure data sequence of the starting monitoring point and the corrected temperature sequence of each monitoring point; Performing weighted fusion processing on the pressure data sequence of the starting monitoring point and the corrected temperature sequence of each monitoring point according to the weight distribution scheme to obtain a comprehensive parameter sequence; The least square method is used to establish a linear correspondence between the oil level change sequence and the comprehensive parameter sequence, so as to obtain a dynamic correlation model between the oil level change inside the transformer bushing and the pressure and temperature.

7. The transformer bushing defect monitoring method based on multidimensional data fusion according to claim 6, characterized in that: The method of performing multidimensional data fusion on the real-time collected pressure data sequence and temperature data sequence according to the dynamic correlation model and the weight distribution scheme, establishing a multidimensional feature vector including pressure, temperature and oil level changes, and performing transformer bushing defect monitoring based on the multidimensional feature vector to obtain monitoring results specifically includes: performing weighted processing on the pressure data sequence of the starting monitoring point and the corrected temperature sequence of each monitoring point according to the weight distribution scheme to obtain a weighted pressure feature of the starting monitoring point and a weighted temperature feature of each monitoring point; Inputting the weighted pressure characteristic of the starting monitoring point and the weighted temperature characteristic of each monitoring point into the dynamic correlation model to obtain the oil level change characteristic; Establishing a standard feature vector based on the weighted pressure feature of the starting monitoring point, the weighted temperature feature of each monitoring point, and the oil level change feature; Based on the structure of the standard feature vector, multi-dimensional data fusion is performed on the real-time collected pressure data sequence and temperature data sequence to establish a multi-dimensional feature vector containing pressure, temperature and oil level changes; Inputting the multidimensional feature vector into a trained defect recognition model to obtain the confidence level of each defect category identified by the multidimensional feature vector; Based on the confidence level of each defect category identified by the multidimensional feature vector, it is determined whether the transformer bushing has a defect type corresponding to the defect category. If so, the severity of the defect type is determined based on the corresponding confidence level, so as to obtain a monitoring result based on the existing defect type and its severity.

8. The transformer bushing defect monitoring method based on multidimensional data fusion according to claim 1, characterized in that: The method further comprises: A defect evolution trend analysis is performed on the defect types included in the monitoring results according to a historical monitoring database, and a corresponding defect early warning mechanism is triggered according to the analysis results and the monitoring results.

9. The transformer bushing defect monitoring method based on multidimensional data fusion according to claim 8, characterized in that: The defect evolution trend analysis of the defect types included in the monitoring results is performed based on the historical monitoring database, and a corresponding defect early warning mechanism is triggered based on the analysis results and the monitoring results, specifically including: Taking the defect type included in the monitoring result as the target defect, extracting historical record data corresponding to defects of the same type as the target defect from the historical monitoring database; wherein the historical record data includes each time point when the defect of the same type was identified and its severity; Perform defect evolution trend analysis based on the historical record data to obtain the average evolution speed of the target defect; Taking the severity of the target defect included in the monitoring result as the current severity, and predicting the deterioration time required for the target defect to reach the maximum severity from the current severity according to the average evolution rate; A corresponding defect warning mechanism is triggered according to the deterioration time and the current severity; wherein the defect warning mechanism is divided into different warning levels based on different deterioration times and different current severity.

10. A transformer bushing defect monitoring system based on multi-dimensional data fusion, characterized in that: include: A multidimensional data matrix establishment module is used to collect pressure data sequences and temperature data sequences of multiple monitoring points inside the transformer bushing according to a preset acquisition frequency, and to establish a multidimensional data matrix containing the monitoring point location, acquisition time point, pressure and temperature based on a sliding window mechanism; a pressure field reconstruction identification module, configured to obtain the pressure change rate of each monitoring point at each acquisition time point based on the multidimensional data matrix, and to obtain the starting monitoring point and starting time point of the pressure field reconstruction when a pressure field reconstruction event is determined to be triggered based on the pressure change rate; a response time difference matrix establishment module, configured to obtain a time delay coefficient of temperature change relative to pressure change at each monitoring point based on the starting time point and the multidimensional data matrix, and establish a response time difference matrix including monitoring point positions and time delay coefficients; a temperature sequence timing correction module, configured to perform timing correction on the temperature data sequence in the multidimensional data matrix based on the starting monitoring point and the response time difference matrix to obtain a correction data matrix; A correlation model and weight acquisition module is used to establish a dynamic correlation model between the oil level change inside the transformer bushing and the pressure and temperature based on the starting monitoring point and the correction data matrix, and determine the weight distribution scheme of the pressure and temperature in the dynamic correlation model; The transformer bushing defect monitoring module is used to perform multidimensional data fusion on the real-time collected pressure data sequence and temperature data sequence according to the dynamic association model and the weight distribution scheme, establish a multidimensional feature vector containing pressure, temperature and oil level changes, and perform transformer bushing defect monitoring based on the multidimensional feature vector to obtain monitoring results; wherein the monitoring results include the defect type and its severity.

Images