A steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility state monitoring

Abstract

The present application relates to power transmission monitoring technical field, disclose a kind of based on power transmission and distribution facility condition monitoring steady-state high-frequency electromagnetic spectrum analysis method, comprising: multivariate acquisition;State abnormality determination;Spectrum feature extraction;Electrical risk assessment;Structural risk positioning;Insulation performance parameter acquisition;Insulation risk positioning;Fault type identification;Early warning generation.The present application is by abnormal preliminary screening as starting point to type identification as terminal progressive analysis logic, comprehensive winding resistance, core magnetic field and insulation oil discharge intensity and other multidimensional features, by capturing the time domain correlation of multiple parameters Collaborative evolution mode, simultaneously analyze the time sequence correlation characteristics of double candidate positions to determine the fault type, realize the diagnosis deepening from risk alarm to fault identification, effectively solve the problem of fault early warning delay and type misjudgment due to single monitoring dimension, static diagnostic model.

Description

Technical Field

[0001] This invention relates to the field of power transmission monitoring technology, and in particular to a steady-state high-frequency electromagnetic spectrum analysis method based on the condition monitoring of power transmission and distribution facilities. Background Technology

[0002] As modern power systems rapidly evolve towards higher reliability and intelligence, the operational stability of oil-immersed transformers, as core nodes of the transmission and distribution network, directly determines regional power supply security and power quality. The strong coupling of multiple physical fields within the transformer, including electrical, magnetic, thermal, and insulation fields, makes its fault mechanisms complex, with early symptoms often subtle and insidious. Traditional operation and maintenance models combining periodic monitoring and manual inspections often fail to capture subtle changes in the early stages of faults, leading to delayed fault warnings and potentially causing equipment damage or even large-scale power outages. Therefore, achieving accurate early warning of transformer conditions is a crucial technological foundation for proactively mitigating systemic risks to the power grid and supporting predictive maintenance strategies.

[0003] Existing technologies include methods for monitoring and analyzing transformer operating status, such as: numbering each transformer in the power grid, monitoring the transformer's operating power parameters in real time, and analyzing the operating status index of a single transformer or the entire power grid; comparing the status index with preset thresholds to screen transformers that may have faults; further analyzing the fault type and generating maintenance information based on the transformer's location; subsequently matching and dispatching maintenance personnel, and managing the maintenance process. However, these existing technologies have the following problems: relying on simple threshold screening of operating power status easily leads to frequent false alarms and missed alarms in fault identification results; relying on single-dimensional analysis of status indices easily leads to delayed perception of early-stage, slowly changing complex faults, failing to achieve early warning of risks in their infancy; relying on fuzzy fault type screening easily leads to gaps in maintenance information, a lack of targeted maintenance preparation, and low maintenance efficiency. Summary of the Invention

[0004] Therefore, this invention provides a steady-state high-frequency electromagnetic spectrum analysis method based on the condition monitoring of power transmission and distribution facilities, which overcomes the problems of delayed fault warning and misjudgment of type caused by the single monitoring dimension and static diagnostic model in the prior art through multi-dimensional fault feature fusion and collaborative diagnosis.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring, comprising: Obtain the winding impedance variation and core flux distribution anomaly values ​​of the oil-immersed transformer; The state of the oil-immersed transformer is determined based on the spatiotemporal relationship between the winding impedance change value and the magnetic flux distribution anomaly value. Based on the abnormal state results of the oil-immersed transformer, the high-frequency transient energy value and spectral distortion value of the steady-state high-frequency electromagnetic spectrum of the oil-immersed transformer are obtained. The existence of electrical risks is determined based on the temporal and spatial characteristics of the winding impedance change value, the high-frequency transient energy value, and the spectral distortion value. Based on the electrical risk assessment results, the first candidate position is selected according to the distribution characteristics of the high-frequency transient energy value and the correlation characteristics of the winding impedance change value, the spectral distortion value and the abnormal magnetic flux distribution value. Obtain the insulation discharge intensity value of the insulating oil in the oil-immersed transformer; The second candidate position is determined based on the relevant characteristics of the spectral distortion value, the magnetic flux distribution anomaly value, the high-frequency transient energy value, and the insulation discharge intensity value. The electrical fault type is determined based on the temporal correlation characteristics of the first candidate position and the second candidate position within a preset time period; A risk warning is generated based on the electrical fault type.

[0006] As a preferred embodiment of the steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility status monitoring described in this invention, the step of determining whether the state of the oil-immersed transformer is abnormal includes: Based on the temporal relationship between the winding impedance change value and the magnetic flux distribution anomaly value, a historical state vector is constructed, and the state mean is determined based on the historical state vector; according to the spatial distance between the historical state vector and the state mean, a historical vector distance set is determined. Based on the temporal relationship between the current winding impedance change value and the magnetic flux distribution anomaly value, a current state vector is constructed, and the current state mean is determined based on the current state vector; the current vector distance is determined based on the spatial distance between the current state vector and the current state mean. The comparison distance is determined based on the spatial distance in the historical vector distance set corresponding to the preset level; Based on the comparison result between the current vector distance and the comparison distance, it is determined that the state of the oil-immersed transformer is abnormal.

[0007] As a preferred embodiment of the steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring described in this invention, the determination of whether an electrical risk exists includes: Based on the winding impedance change value, the high-frequency transient energy value, and the spectral distortion value at each moment within the next preset abnormal time period, time series sequences are constructed respectively to obtain impedance time series sequence, energy time series sequence, and distortion time series sequence. Based on the linear trends of the impedance time series, the energy time series, and the distortion time series, the impedance strength coefficient, energy strength coefficient, and distortion strength coefficient are quantitatively determined. The condition of simultaneous upward movement is determined based on the consistency of the signs of the impedance strength coefficient, energy strength coefficient, and distortion strength coefficient. Based on the aforementioned upward trend condition, impedance rate change, energy rate change, and distortion rate change are determined according to the respective rates of change of the impedance time series, the energy time series, and the distortion time series at adjacent moments within the preset abnormal duration. The existence of electrical risks is determined based on the impedance speed change, the energy speed change, and the distortion speed change.

[0008] As a preferred embodiment of the steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring described in this invention, the determination of whether there is an electrical risk based on the impedance speed change, the energy speed change, and the distortion speed change includes: Calculate the ratio of the number of times during which the impedance rate change, the energy rate change, and the distortion rate change are all positive within the preset abnormal duration to the total number of times during the preset abnormal duration, in order to obtain the synchronization rate of change. When the synchronous change rate is greater than a preset change threshold, an electrical risk is determined to exist.

[0009] As a preferred embodiment of the steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring described in this invention, the screening of the first candidate positions includes: The cross-correlation degree is determined based on the cross-correlation relationship of the time series of the high-frequency transient energy values ​​at different preset spatial locations, and the spatial correlation is determined based on the threshold screening results of the cross-correlation degree. The preset spatial point corresponding to the maximum value among the high-voltage transient energy, low-voltage transient energy, and iron core transient energy determined based on the time-domain energy characteristics of the high-frequency transient energy values ​​at different preset spatial points is marked as the energy dominance point; The time sequence of recording the high-frequency transient energy values ​​of the energy-dominant point within the preset positioning time is the dominant energy sequence; The first candidate position is selected based on the spatial correlation and the correlation characteristics between the dominant energy sequence and the winding impedance change value, the spectral distortion value, and the magnetic flux distribution anomaly value.

[0010] As a preferred embodiment of the steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring described in this invention, the method for determining cross-correlation degree based on the cross-correlation relationship of the time series of the high-frequency transient energy values ​​at different preset spatial locations, and determining spatial correlation based on the threshold screening results of the cross-correlation degree, includes: Based on the temporal characteristics of the high-frequency transient energy values ​​at the preset spatial locations, high-voltage energy sequences, low-voltage energy sequences, and core energy sequences are constructed respectively. The correlation between high and low pressure, high-pressure and low-pressure, and low-pressure are determined based on the correlation characteristics between the high-pressure energy sequence, the low-pressure energy sequence, and the core energy sequence, respectively. The spatial correlation is determined to be a strong spatial correlation based on the comparison results of the high-low pressure cross-correlation, the high-speed rail pressure cross-correlation, and the low-speed rail cross-correlation with the same threshold.

[0011] As a preferred embodiment of the steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring described in this invention, the step of screening the first candidate position based on the spatial correlation and the correlation characteristics of the dominant energy sequence, the winding impedance change value, the spectral distortion value, and the magnetic flux distribution anomaly value includes: Based on the time-series characteristics of the winding impedance change value, the spectral distortion value, and the magnetic flux distribution anomaly value, a winding resistance sequence, a harmonic distortion sequence, and a magnetic field strength sequence are constructed. The resistance oscillation correlation degree and the magnetic field oscillation correlation degree are determined according to the correlation degree between the dominant energy sequence and the winding resistance sequence and the magnetic field strength sequence, respectively. The wave resistance synchronization ratio is determined based on the cooperative characteristics of the changing trends of the winding resistance sequence and the harmonic distortion sequence within the preset positioning time. The harmonic distortion variance is determined based on the statistical fluctuation characteristics of the harmonic distortion sequence. Based on the strong spatial correlation, and according to the multi-threshold joint determination results of the resistance oscillation correlation, the magnetic field oscillation correlation, and the wave impedance synchronization ratio, the energy dominance point is determined as the first candidate position.

[0012] As a preferred embodiment of the steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring described in this invention, the determination of the second candidate location includes: The magnetic resonance synergy index is determined based on the correlation between the local intensity sequence and the transient amplitude sequence constructed from the normalized temporal characteristics of the magnetic flux distribution anomaly and the high-frequency transient energy value, respectively. The rate of change of intensity is determined based on the degree of difference between the insulation discharge intensity value and the preset intensity threshold. The distortion rate is determined based on the degree of difference between the spectral distortion value and the preset harmonic threshold. The second candidate position is determined based on the threshold comparison results of the magnetic resonance synergy index, the intensity change rate, and the distortion change rate.

[0013] As a preferred embodiment of the steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring described in this invention, the method includes: determining the second candidate position based on the threshold comparison results of the magnetic resonance coordination index, the intensity change rate, and the distortion change rate, comprising: Based on the threshold screening results of the magnetic resonance synergy index, the risk type is determined to be strong magnetic circuit related type; Based on the threshold screening results of the magnetic resonance synergy index, the intensity change rate, and the distortion change rate, the second candidate position is determined to be an oil-paper composite insulation system.

[0014] As a preferred embodiment of the steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring described in this invention, the step of determining the electrical fault type based on the temporal correlation characteristics of the first candidate position and the second candidate position within a preset time period includes: The positional coordination ratio is determined statistically based on the coordination time characteristics of the first and second candidate positions. When the position coordination ratio is greater than a preset ratio threshold and the first candidate position is the energy dominance point determined based on the high voltage transient energy, the electrical fault type is determined to be a winding short circuit fault. When the position coordination ratio is greater than a preset ratio threshold and the first candidate position is the energy dominance point determined based on the low-voltage transient energy, the electrical fault type is determined to be a winding overload fault. When the positional coordination ratio is greater than a preset ratio threshold, and the first candidate position is the energy dominance point determined based on the transient energy of the iron core, the electrical fault type is determined to be an iron core flux leakage fault.

[0015] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention effectively solves the problems of delayed fault warning and misclassification caused by the single monitoring dimension and static diagnostic model in existing technologies by constructing a multi-dimensional, progressive steady-state high-frequency electromagnetic spectrum analysis framework. This invention simultaneously acquires two heterogeneous physical quantities, winding impedance and core flux, and performs initial anomaly screening based on their spatiotemporal characteristics, overcoming the limitations of single-parameter monitoring. Furthermore, it adaptively distinguishes between normal operating condition fluctuations and early weak anomalies using Mahalanobis distance and dynamic thresholds, significantly improving detection sensitivity and reliability. Secondly, based on the initial anomaly screening, high-frequency transient energy and spectral distortion parameters are introduced. By capturing the collaborative evolution patterns of multiple parameters in the time domain, it effectively distinguishes between instantaneous interference and continuously deteriorating fault precursors, achieving a leap from anomaly detection to risk situation assessment. Further, this invention designs parallel first and second candidate location screening mechanisms, utilizing spatial cross-correlation, energy-dominant point discrimination, multi-parameter correlation verification, and indicators such as magnetic resonance synergy index and intensity change rate to accurately locate the fault source from two independent dimensions, avoiding misjudgment and missed detections along a single path. This invention analyzes the temporal correlation characteristics of two candidate locations within a preset time period and combines them with the spatial attributes of the energy-dominant point to accurately determine specific electrical fault types such as winding short circuits, winding overloads, or core flux leakage. This enables a deeper diagnostic process from risk alarm to fault identification, providing clear and direct classification criteria for operation and maintenance decisions, thereby significantly improving the intelligence level and early warning accuracy of power transmission and distribution facility status monitoring. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the overall process logic of a steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility status monitoring, provided as an embodiment of the present invention.

[0018] Figure 2 The present invention provides a logic diagram for determining the abnormal state of an oil-immersed transformer using a steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility status monitoring, as an embodiment of the present invention.

[0019] Figure 3 The present invention provides a logic diagram for determining the existence of electrical risks using a steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility status monitoring, as an embodiment of the present invention.

[0020] Figure 4The present invention provides a logic diagram for determining strong spatial correlation in a steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility status monitoring, as an embodiment of the present invention. Detailed Implementation

[0021] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0022] Example 1, referring to Figure 1 As one embodiment of the present invention, a steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring is provided, such as... Figure 1 The specific steps shown are as follows: S100: Obtain the winding impedance change value and the magnetic flux distribution anomaly value of the core of the oil-immersed transformer. S200: Determine whether the state of the oil-immersed transformer is abnormal based on the spatiotemporal relationship between the winding impedance change value and the abnormal magnetic flux distribution value. S300: Based on the abnormal state results of the oil-immersed transformer, obtain the high-frequency transient energy value and spectral distortion value of the steady-state high-frequency electromagnetic spectrum of the oil-immersed transformer. S400: Determine whether there is an electrical risk based on the temporal and spatial characteristics of winding impedance change, high-frequency transient energy value, and spectral distortion value; S500: Based on the electrical risk assessment results, the first candidate position is selected according to the distribution characteristics of high-frequency transient energy values ​​and the correlation characteristics of winding impedance changes, spectral distortion values ​​and abnormal magnetic flux distribution values. S600: Obtain the insulation discharge intensity value of the insulating oil in an oil-immersed transformer; S700: The second candidate position is determined based on the relevant characteristics of spectral distortion value, magnetic flux distribution anomaly value, high-frequency transient energy value, and insulation discharge intensity value. S800: Determine the electrical fault type based on the temporal correlation characteristics of the first and second candidate positions within a preset time period; S900: Generates risk warnings based on electrical fault types.

[0023] In this embodiment of the invention, the steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility status monitoring is applied to the health status monitoring of power transmission and distribution equipment such as transformers. By analyzing the high-frequency electromagnetic spectrum during equipment operation, electrical faults and anomalies caused by overload, local short circuit, insulation aging, etc. are captured in real time, providing early risk analysis for management decision-making reference, thereby improving the safety and reliability of the equipment and reducing the risk of equipment failure and downtime.

[0024] In this embodiment of the invention, a multi-dimensional and multi-level state assessment and risk analysis method is constructed by acquiring multi-dimensional parameters. The oil-immersed transformer is in normal energized operation. Through online monitoring units installed on and around the transformer body, various electromagnetic and electrical signals reflecting the transformer's operating status are collected, and the corresponding operating status values ​​are further extracted. Specifically, the winding impedance change value characterizes the change in the electrical characteristics of the winding, including resistance and inductive reactance, relative to the normal state at a specific AC frequency. It can be based on high-frequency components in the operating current and voltage signals, or by processing the response characteristics of applied disturbance signals, to extract the equivalent impedance change of the winding within a preset frequency band, and its amplitude change or phase shift is used as the winding impedance change value. The magnetic flux distribution anomaly value analyzes the uniformity of magnetic flux density at different locations in the core. Multiple magnetic field detection units can be deployed on the core surface or in the adjacent area to collect magnetic flux density signals at corresponding locations, and the degree of unevenness in magnetic flux distribution is calculated based on the differences between different measurement points. The high-frequency transient energy value and spectral distortion value are obtained by analyzing the high-frequency electromagnetic signals generated during transformer operation. These high-frequency electromagnetic signals can be obtained by installing monitoring units on the transformer body. The high-frequency transient energy value is obtained from nearby electromagnetic induction devices or signal acquisition units. It is used to measure the electromagnetic wave energy radiated by rapid transient events such as partial discharge and electric arc in a specific high-frequency band, such as 300MHz-3GHz. It can be obtained by envelope analysis or energy statistics of the acquired signal within a preset time window. The spectral distortion value is a comprehensive index used to measure the waveform distortion or spectral purity of high-frequency electromagnetic signals. It can be obtained by performing frequency domain analysis on high-frequency electromagnetic signals and calculating the proportion or distribution dispersion of non-dominant frequency components in the overall spectrum. The insulation discharge intensity value is a comprehensive parameter used to directly characterize the intensity of partial discharge activity inside the oil-paper insulation system. It can be based on the collection of discharge pulses, electromagnetic radiation or ultrasonic signals generated in the insulating oil by the partial discharge detection device, and form an insulation discharge intensity value to characterize the change of insulation state according to its amplitude, frequency or energy characteristics.

[0025] In this embodiment of the invention, by designing two parallel logical branches—a first candidate location and a second candidate location—the aim is to more accurately identify potential electrical faults in power transmission and distribution facilities, especially under complex multi-physical processes. Electrical equipment faults are typically the result of multiple factors and hierarchical interplay; a single fault mode cannot fully reveal the equipment's health status. Specifically, winding overload, partial short circuit, and core faults can all occur simultaneously and influence each other, leading to different types of fault signals at different locations. Therefore, designing two parallel candidate location logical branches allows the system to analyze in parallel at multiple monitoring points, identifying different types of fault sources, thus avoiding the limitations of traditional methods that rely on a single monitoring path. The first candidate location, by analyzing the transient oscillation signals of high voltage, low voltage, and the core, combined with multi-parameter characteristics such as resistance changes, magnetic field strength, and high-frequency harmonic distortion, can identify faults such as partial short circuits, winding overloads, and core problems. The second candidate location focuses on the oil-paper composite insulation system, independently assessing the degradation of the insulation system by monitoring parameters such as changes in the dielectric strength of the insulating oil, oscillation amplitude, and harmonic distortion. This dual monitoring path not only enables comprehensive assessment from multiple physical locations but also, through parallel analysis of different fault mechanisms, further reduces false alarms and missed alarms caused by environmental changes or single parameters, thereby improving the accuracy and reliability of fault diagnosis. Furthermore, by analyzing the temporal correlation between these two candidate locations, the fault source can be identified more precisely. For example, when the oscillation amplitude on the high-voltage side changes synchronously with the degradation of the insulation system, it can be determined that a high-voltage fault may lead to a decline in the performance of the insulating oil, thus diagnosing potential risks to the equipment at an early stage and avoiding latent faults that traditional methods cannot detect in a timely manner.

[0026] The preset time limit is the length of the time window used to analyze the temporal correlation features of the first and second candidate locations. It depends on the sampling frequency of the monitoring system and the speed of fault feature evolution. It is usually set between 30 minutes and 2 hours. In this embodiment, it is set to 1 hour to ensure that the dynamic correlation features in the development process of electrical faults are fully captured.

[0027] By simultaneously acquiring two heterogeneous physical quantities—winding impedance and core flux—and performing initial anomaly screening based on their spatiotemporal characteristics, the limitations of single-parameter monitoring are overcome. Building upon this initial screening, high-frequency transient energy and spectral distortion parameters are introduced. By capturing the collaborative evolution patterns of multiple parameters in the time domain and their correlations, a leap from anomaly detection to risk assessment is achieved, effectively distinguishing between transient interference and continuously deteriorating fault precursors. Furthermore, by designing parallel first and second candidate location screening mechanisms, fault source localization is achieved from two independent dimensions: electrical-magnetic circuit coupling characteristics and insulation-spectral correlation characteristics. The fault type is determined by analyzing the temporal correlation characteristics of the two candidate locations, achieving a deeper diagnostic process from risk alarm to fault identification. This provides a direct and clear classification basis for operation and maintenance decisions. Through multi-source information fusion and spatiotemporal correlation analysis, the problems of delayed fault warnings and misclassification due to single monitoring dimensions and static diagnostic models are effectively solved.

[0028] Example 2, refer to Figures 2-4 Based on the previous embodiment, this embodiment provides a specific implementation method for steady-state high-frequency electromagnetic spectrum analysis based on power transmission and distribution facility status monitoring, in order to illustrate the technical means of this method.

[0029] Please see Figure 2 As shown, this is the logic diagram for determining the abnormal state of an oil-immersed transformer in this embodiment. In this embodiment, the process of determining whether the state of the oil-immersed transformer is abnormal based on the spatiotemporal characteristic relationship between the winding impedance change value and the abnormal magnetic flux distribution value includes: Based on the temporal relationship between the winding impedance change value and the magnetic flux distribution anomaly value, a historical state vector is constructed, and the state mean is determined based on the historical state vector; the historical vector distance set is determined according to the spatial distance between the historical state vector and the state mean. Based on the temporal relationship between the current winding impedance change value and the magnetic flux distribution anomaly value, a current state vector is constructed, and the current state mean is determined based on the current state vector; the current vector distance is determined based on the spatial distance between the current state vector and the current state mean. The comparison distance is determined based on the spatial distance in the historical vector distance set corresponding to the preset level; When the current vector distance is greater than the comparison distance, the state of the oil-immersed transformer is determined to be abnormal.

[0030] In this embodiment of the invention, the preset distribution duration is the length of the time window used to construct the historical state vector. It depends on the sampling frequency of the monitoring system and the urgency of the transformer state changes, and is usually set between 1 hour and 24 hours. In this embodiment, it is set to 4 hours to ensure that enough historical data is collected to establish a stable normal state statistical model. The preset historical quantile value is the quantile used to extract the threshold from the ascending sequence of historical Mahalanobis distances. It depends on the sensitivity requirements of anomaly detection and the control of false alarm rate, and is usually set between 0.90 and 0.99. In this embodiment, it is set to 0.95, which can effectively distinguish between normal state fluctuations and abnormal state deviations.

[0031] It should be noted that by constructing a multivariate state space using historical data of winding resistance and magnetic field strength within a preset distribution period, and calculating the Mahalanobis distance between the current state vector and the historical normal distribution mean, not only are the amplitude changes of each parameter considered, but the inherent correlation and covariance characteristics between them are also captured. By selecting the distance corresponding to the preset historical quantile value as a dynamic threshold, it can adaptively follow the natural fluctuations of the equipment's normal operating state, effectively avoiding false alarms caused by normal operating condition fluctuations such as load changes and ambient temperature fluctuations. At the same time, it significantly improves the detection sensitivity and reliability of early, weak, and correlated latent anomalies, such as slight magnetic circuit saturation accompanied by local overheating.

[0032] In this embodiment of the invention, the process of determining whether there is an electrical risk based on the temporal spatial characteristics of winding impedance change value, high-frequency transient energy value, and spectral distortion value includes: Based on the winding impedance change value, high-frequency transient energy value and spectral distortion value at each moment within the next preset abnormal time period, time series sequences are constructed respectively to obtain impedance time series sequence, energy time series sequence and distortion time series sequence. The linear fitting slope of the impedance time series is calculated to obtain the impedance strength coefficient, the linear fitting slope of the energy time series is calculated to obtain the energy strength coefficient, and the linear fitting slope of the distortion time series is calculated to obtain the distortion strength coefficient. When the impedance strength coefficient, energy strength coefficient, and distortion strength coefficient are all positive, it is determined that the condition of rising in the same direction is met. Based on the condition of rising in the same direction, the rate of change of the impedance time series at adjacent moments within a preset abnormal duration is calculated to obtain several impedance rate changes, and the rate of change of the energy time series at adjacent moments within a preset abnormal duration is calculated to obtain several energy rate changes, and the rate of change of the distorted time series at adjacent moments within a preset abnormal duration is calculated to obtain several distortion rate changes. The existence of electrical risks is determined based on impedance speed change, energy speed change, and distortion speed change.

[0033] Specifically, the preset anomaly duration is the length of the time window used for preliminary dynamic analysis of electrical risks. It depends on the speed of fault development and the real-time requirements of the system, and is usually set between 30 minutes and 2 hours. In this embodiment, it is set to 1 hour, which can effectively balance the timeliness of fault warning and data stability.

[0034] It should be noted that the fixed-length window effectively isolates transient interference signals, providing a high-quality data foundation for the subsequent extraction of robust time-series evolution features. Based on this, the intensity coefficients of each sequence are calculated through linear fitting, quantifying the overall change trend of each parameter within the window into a clear slope index. This achieves an objective representation of the parameter evolution direction. By setting a condition of simultaneous upward movement as a judgment criterion, the consistent degradation trend of multiple key fault symptom parameters can be captured simultaneously, enhancing the early identification capability of complex faults or systemic defects. Furthermore, under the premise of satisfying the condition of simultaneous upward movement, the dynamic characteristics of parameter changes are revealed by calculating the rate of change of each parameter at adjacent time points, i.e., whether its degradation process is in a linear accumulation stage or has entered an accelerated development stage. Based on multi-dimensional dynamic characteristics, a comprehensive risk assessment is performed, achieving early warning of potential faults and providing a key basis for distinguishing risk levels for predictive maintenance decisions.

[0035] For further details, please refer to Figure 3 As shown, this is the logic diagram for determining the existence of electrical risks in this embodiment. In this embodiment, the process of determining whether there is an electrical risk based on impedance speed change, energy speed change, and distortion speed change includes: Calculate the ratio of the number of times during the preset abnormal time when the impedance change rate, energy change rate, and distortion change rate are all positive to the total number of times during the preset abnormal time, in order to obtain the synchronization rate of change. An electrical risk is identified when the rate of synchronous change exceeds a preset change threshold.

[0036] Specifically, the preset change threshold is a benchmark value used to judge whether the consistency of the change rate direction of the three parameters of impedance, energy and distortion reaches the risk level. It depends on the statistical distribution characteristics of the synchronous change rate of the system under normal conditions and the balance requirements of fault detection sensitivity and false alarm rate control. It is usually set between 0.70 and 0.90. In this embodiment, it is set to 0.75, which can effectively control the false alarm rate within an acceptable range while ensuring high detection sensitivity for early potential electrical faults.

[0037] It should be noted that by calculating the synchronous change rate, the analysis dimension is elevated from the absolute value or overall trend of the parameter to the level of the synergy of its instantaneous change rate. This can identify the inherent linkage of multiple key characteristic quantities in the evolution process. By using synchronous measurement based on the consistency of change direction and quantifying thresholds to distinguish between normal random fluctuations and abnormal co-evolution patterns, it can effectively characterize the unified and concurrent impact dynamics of potential failure mechanisms on different aspects of the system, reducing the possibility of misjudgment triggered by accidental fluctuations or brief disturbances of a single parameter.

[0038] Furthermore, in this embodiment of the invention, the process of selecting the first candidate position based on the distribution characteristics of high-frequency transient energy values ​​and the correlation characteristics of winding impedance changes, spectral distortion values, and magnetic flux distribution anomalies includes: The cross-correlation degree is determined based on the cross-correlation relationship of time series of high-frequency transient energy values ​​at different preset spatial locations, and spatial correlation is determined based on the threshold screening results of the cross-correlation degree. The sum of squares of the high-frequency transient energy values ​​at each moment within the preset positioning time for three different preset spatial points is calculated to obtain the high-voltage transient energy, low-voltage transient energy and iron core transient energy respectively, and the preset spatial point corresponding to the maximum value among the high-voltage transient energy, low-voltage transient energy and iron core transient energy is marked as the energy dominance point; The time sequence of recording the high-frequency transient energy values ​​of the energy-dominant point within a preset positioning time is the dominant energy sequence; The first candidate position was selected based on spatial correlation and the correlation characteristics of dominant energy sequence, winding impedance change value, spectral distortion value, and magnetic flux distribution anomaly value.

[0039] In this embodiment of the invention, the preset spatial points include high-voltage side monitoring points, low-voltage side monitoring points, and core monitoring points. These are located on the exposed metal surfaces of the transformer's high-voltage bushing riser or corresponding tank wall, low-voltage bushing riser or corresponding tank wall, and core clamps, respectively. These three points constitute a key monitoring network covering the transformer's core electrical and magnetic circuit structure: the high-voltage and low-voltage side points are designed to directly capture ultra-high frequency electromagnetic radiation generated by faults such as partial discharges originating from their respective winding circuits; the core point is used to sense abnormal leakage magnetic fields and their modulation signals generated by magnetic circuit faults (such as inter-laminar short circuits).

[0040] In this embodiment of the invention, the preset positioning duration is the length of the time window used for spatial correlation analysis of fault sources and identification of energy dominance points. It depends on the spatial propagation characteristics of the fault signal and the real-time requirements of the system positioning. It is usually set between 15 minutes and 1 hour. In this embodiment, it is set to 30 minutes, which can ensure that enough spatial synchronization data is collected to reliably characterize the correlation and energy distribution characteristics of oscillation signals at different points.

[0041] It should be noted that by simultaneously analyzing the cross-correlation of transient oscillation signals at three preset spatial points, it is possible to preliminarily determine whether the fault belongs to a local strong source or system coupling based on the spatial distribution pattern, effectively filtering interference signals generated by electromagnetic propagation or shared loops. Based on this, by calculating and comparing the cumulative oscillation energy at each point within the same time period, the physical location where energy release is most concentrated can be objectively identified, avoiding positioning errors caused by signal attenuation or reflection within complex structures. Simultaneously, by verifying the multi-parameter correlation between the oscillation sequence of the energy-dominant point and winding resistance, harmonic distortion, and magnetic field strength, it is possible to confirm whether the electromagnetic activity at this location is statistically coordinated with the abnormal evolution of other key state parameters, ensuring that the positioning results conform to the physical mechanism of the fault and achieving highly reliable and interpretable fault source positioning.

[0042] For further details, please refer to Figure 4 As shown, this is the logic diagram for determining strong spatial correlation in this embodiment. In this embodiment, the process of determining the cross-correlation degree based on the cross-correlation relationship of the time series of high-frequency transient energy values ​​of different preset spatial points, and determining the spatial correlation based on the threshold screening results of the cross-correlation degree includes: High-voltage energy sequence, low-voltage energy sequence and iron core energy sequence are constructed based on the high-frequency transient energy values ​​at each time within a preset positioning time at three different preset spatial locations. Calculate the Pearson correlation coefficient between the high-pressure energy sequence and the low-pressure energy sequence to obtain the high-low pressure cross-correlation, and calculate the Pearson correlation coefficient between the high-pressure energy sequence and the core energy sequence to obtain the high-pressure cross-correlation, and calculate the Pearson correlation coefficient between the low-pressure energy sequence and the core energy sequence to obtain the low-pressure cross-correlation. When the correlation between high and low voltage is greater than a preset correlation threshold, the correlation between high-speed rail and low-voltage is greater than a preset correlation threshold, and the correlation between low-speed rail and low-voltage is greater than a preset correlation threshold, the spatial correlation is determined to be a strong spatial correlation.

[0043] Otherwise, the spatial association is determined to be a weak spatial association.

[0044] Specifically, the preset correlation threshold is the lower limit of the Pearson correlation coefficient used to determine whether the transient energies at different monitoring points have a strong correlation. It depends on the system's requirements for the spatial consistency of fault signals and is usually set between 0.6 and 0.9. In this embodiment, it is set to 0.8, which can effectively distinguish between global oscillations and local independent interference caused by the same fault source.

[0045] It should be noted that by synchronously analyzing the correlation between the pairwise high-frequency transient energy values ​​of three key monitoring points to determine whether there is a strong spatial correlation between signals, an effective filter can be constructed directly from the spatial consistency characteristics of the data to distinguish disturbances of different natures. By distinguishing disturbances, oscillation signals that exhibit high synchronicity at multiple independent monitoring points can be identified, characterizing systemic faults originating from the transformer's internal main circuit or magnetic circuit. At the same time, it can effectively filter out isolated interferences or localized micro-discharges that occur only at a single point and lack statistical correlation with each other, improving the spatial specificity and anti-interference capability of preliminary fault location.

[0046] Furthermore, in this embodiment of the invention, the process of selecting the first candidate location based on spatial correlation and the correlation characteristics of the dominant energy sequence and winding impedance change value, spectral distortion value, and magnetic flux distribution anomaly value includes: Based on the winding impedance change value, spectral distortion value, and magnetic flux distribution anomaly value at each moment within the preset positioning time, the winding resistance sequence, harmonic distortion sequence, and magnetic field strength sequence are constructed respectively. The Pearson coefficients of the normalized dominant energy sequence and the normalized winding resistance sequence are calculated to obtain the resistance oscillation correlation. The Pearson coefficients of the normalized dominant energy sequence and the normalized magnetic field strength sequence are also calculated to obtain the magnetic field oscillation correlation. The proportion of times when the rate of change of the winding resistance sequence and the rate of change of the harmonic distortion sequence have the same sign between two adjacent moments within a preset positioning time is used to obtain the wave resistance synchronization ratio. Calculate the variance of the harmonic distortion sequence to obtain the harmonic distortion variance; If the spatial correlation is strong, and at least one of the resistance oscillation correlation and the magnetic field oscillation correlation is greater than the preset oscillation correlation threshold, and the wave resistance synchronization ratio is greater than the preset ratio threshold, then the energy dominance point is determined as the first candidate location.

[0047] Specifically, the preset oscillation correlation threshold is the lower limit of the Pearson coefficient used to determine whether the dominant energy sequence and the resistance / magnetic field sequence have a significant statistical correlation. It depends on the requirements for identifying the fault coupling strength and the signal noise level, and is usually set between 0.5 and 0.8. In this embodiment, it is set to 0.6, which can effectively identify multi-parameter coordinated oscillations caused by the physical process of the fault. The preset ratio threshold is the minimum ratio requirement for determining whether the resistance change rate and the spectral distortion change rate keep changing synchronously. It depends on the judgment criteria for the consistency of the fault development process, and is usually set between 0.6 and 0.9. In this embodiment, it is set to 0.7, which can confirm that the resistance change caused by the fault and the high-frequency spectral distortion have statistical consistency in the evolution trend.

[0048] It should be noted that by performing Pearson correlation analysis on the dominant energy sequence with the winding impedance sequence and magnetic field strength sequence, respectively, the correlation between resistance oscillation and magnetic field oscillation was calculated. Statistical methods were used to objectively verify whether the abnormal electromagnetic energy release was significantly related to the two core fault physical processes of winding heating and core magnetic saturation, providing a data-driven physical explanation for localization. Secondly, by analyzing whether the trends of winding impedance change and spectral distortion change are in the same direction most of the time, a synergistic relationship between circuit condition deterioration and high-frequency spectral pollution was confirmed from a dynamic evolution perspective. Furthermore, calculating the variance of the harmonic distortion sequence helps assess the severity of the spectral disturbance itself, and multi-parameter threshold screening avoids misjudgments that might be caused by a single electromagnetic parameter anomaly.

[0049] In this embodiment of the invention, the process of determining the second candidate location based on the relevant characteristics of spectral distortion value, magnetic flux distribution anomaly value, high-frequency transient energy value, and insulation discharge intensity value includes: Based on the temporal characteristics of the magnetic flux distribution anomalies and high-frequency transient energy values ​​within the preset risk period, local intensity sequences and transient amplitude sequences are constructed respectively. The absolute values ​​of the Pearson correlation coefficients of the normalized local intensity sequences and normalized transient amplitude sequences are calculated to obtain the magnetic resonance synergy index. Calculate the relative deviation between the insulation discharge intensity value and the preset intensity threshold to obtain the intensity change rate; Calculate the relative deviation between the spectral distortion value and the preset harmonic threshold to obtain the distortion change rate; The second candidate position was determined based on the threshold comparison results of the magnetic resonance synergy index, intensity change rate, and distortion change rate.

[0050] Specifically, the preset risk duration is the length of the analysis time window used to assess magnetic resonance synergy and changes in insulation state. It depends on the speed of insulation fault development and the real-time monitoring requirements, and is usually set between 30 minutes and 2 hours. In this embodiment, it is set to 1 hour, which can effectively capture the dynamic correlation in the evolution of fault characteristics. The preset strength threshold is a reference benchmark value for assessing whether the dielectric strength of insulating oil has significantly deteriorated. It depends on the oil type, operating years, and maintenance standards, and is usually set between 35kV / 2.5mm and 45kV / 2.5mm. In this embodiment, it is set to 40kV / 2.5mm, which can be used to quantify the relative degree of decline in oil insulation performance. The preset harmonic threshold is a reference benchmark value for determining whether high-frequency electromagnetic harmonic distortion is abnormal. It depends on the equipment type, monitoring frequency band, and historical background level, and is usually set between 5% and 10%. In this embodiment, it is set to 8%, which can effectively identify the aggravation of spectral pollution caused by active defects such as partial discharge.

[0051] In this embodiment, the normalization of the normalized local intensity sequence and the normalized transient amplitude sequence adopts the maximum-minimum value normalization process, thereby uniformly mapping the data to the [0,1] interval, which can effectively eliminate the influence of different units on the comparability of data, while preserving the relative distribution characteristics of the sequence data, ensuring that the local intensity sequence and the transient amplitude sequence have equal weight and operability in the subsequent feature fusion analysis.

[0052] It should be noted that calculating the magnetic resonance symmetry index can objectively quantify the statistical correlation strength between the core leakage magnetic field and the transient electromagnetic oscillations caused by partial discharge, thus providing a clear quantitative basis for distinguishing between correlated discharges caused by magnetic circuit faults and general insulation discharges. Secondly, by calculating the rate of change of insulation discharge intensity and spectral distortion values ​​relative to their respective preset benchmark thresholds, the focus of judgment shifts from the absolute level of the parameters to their relative trend of deviation from the normal benchmark, enabling more sensitive detection of the declining trend of insulation dielectric properties and the development trend of high-frequency spectral pollution. Furthermore, by comparing multiple parameter thresholds and determining the second candidate position based on the combined logic of the comparison results, faults in oil-paper insulation systems can be systematically identified, improving the specificity of fault classification and the reliability of early warning.

[0053] Furthermore, the process of determining the second candidate location based on the threshold comparison results of the magnetic resonance symmetry index, intensity change rate, and distortion change rate includes: When the magnetic resonance coordination index is greater than the preset coordination threshold, the risk type is determined to be strong magnetic circuit correlation type; Based on the magnetic resonance coordination index being less than or equal to a preset coordination threshold, when the intensity change rate is less than a preset intensity change threshold and the distortion change rate is greater than a preset distortion threshold, the second candidate position is determined to be the oil-paper composite insulation system.

[0054] Specifically, the preset coordination threshold is the lower limit of the correlation coefficient between the magnetic field strength and the high-frequency transient energy value to determine whether they have strong coordination. It depends on the identification requirements of the coupling strength between the magnetic circuit and the discharge activity, and is usually set between 0.7 and 0.9. In this embodiment, it is set to 0.8, which can effectively distinguish between strongly correlated discharges directly caused by core faults and general insulating discharges. The preset strong change threshold is the relative change rate threshold for determining whether the dielectric strength of the insulating oil has undergone "significant" degradation. It is usually a negative value, depending on the early warning sensitivity requirements for the slow aging of the oil. It is usually set between -0.15 and -0.05. In this embodiment, it is set to -0.10, that is, a decrease of 10%, which can capture the meaningful downward trend of the oil's insulating performance while ignoring small normal fluctuations. The preset distortion threshold is the relative change rate threshold for determining whether the high-frequency electromagnetic harmonic distortion has been significantly aggravated. It is usually a positive value, depending on the early warning requirements for the aggravation of spectral pollution. It is usually set between 0.1 and 0.3. In this embodiment, it is set to 0.20, that is, an increase of 20%, which can effectively identify the substantial increase in harmonic levels caused by the enhancement of partial discharge activity.

[0055] It should be noted that by comparing the magnetic resonance coordination index with a preset coordination threshold, and directly classifying it as a strong magnetic circuit correlation type when it exceeds the threshold, specific risks caused by faults such as core magnetic circuit saturation and inter-laminar short circuits, which have a high statistical correlation between electromagnetic oscillations and leakage magnetic fields, can be directly identified. This avoids confusing such faults with insulation discharge and improves the directionality of initial diagnosis. Secondly, by requiring that the trend of decreasing insulation performance and the trend of increasing high-frequency spectrum pollution must occur simultaneously and significantly, a strict dual-trend consistency criterion is formed. This allows for the specific identification of fault modes dominated by the gradual degradation of the insulation medium and accompanied by partial discharge activity, effectively eliminating misjudgments caused by single parameter fluctuations or other types of interference.

[0056] Furthermore, in this embodiment of the invention, the process of determining the electrical fault type based on the temporal correlation characteristics of the first candidate position and the second candidate position within a preset time period includes: The ratio of the number of times the first and second candidate positions appear simultaneously within a preset time period to the total number of times within the preset time period is used to obtain the position coordination ratio. When the position coordination ratio is greater than the preset ratio threshold and the first candidate position is the energy dominance point determined based on the preset spatial point corresponding to the high voltage transient energy, the electrical fault type is determined to be a winding short circuit fault. When the position coordination ratio is greater than the preset ratio threshold and the first candidate position is the energy dominance point determined based on the preset spatial point corresponding to the low-voltage transient energy, the electrical fault type is determined to be a winding overload fault. When the position coordination ratio is greater than the preset ratio threshold and the first candidate position is the energy dominance point determined based on the preset spatial point corresponding to the transient energy of the iron core, the electrical fault type is determined to be an iron core flux leakage fault.

[0057] Specifically, the preset ratio threshold is the minimum ratio requirement for determining whether the first candidate position and the second candidate position have a stable and significant correlation in time sequence. It depends on the strictness of the judgment on the persistence of fault coupling relationship and is usually set between 0.6 and 0.9. In this embodiment, it is set to 0.7, which can effectively confirm that there is a stable synergistic relationship between structural faults and insulation faults in time development.

[0058] It should be noted that by calculating the proportion of times when the first and second candidate locations appear simultaneously within a preset time period, the stable coexistence relationship between structural and insulation faults over time is identified and verified, effectively improving the statistical confidence and anti-interference capability of type determination. Based on this, the final determination is made by combining the specific spatial attributes of the first candidate locations that have reached the coordination ratio threshold. This logically integrates spatially oriented information with clear physical directionality and independent evidence of insulation risk, distinguishing fault types with similar symptoms but different mechanisms, thus providing a highly specific diagnostic basis for implementing precise maintenance.

[0059] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the present invention.

Claims

1. A steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring, characterized in that, include: Obtain the winding impedance variation and core flux distribution anomaly values ​​of the oil-immersed transformer; The state of the oil-immersed transformer is determined based on the spatiotemporal relationship between the winding impedance change value and the magnetic flux distribution anomaly value. Based on the abnormal state results of the oil-immersed transformer, the high-frequency transient energy value and spectral distortion value of the steady-state high-frequency electromagnetic spectrum of the oil-immersed transformer are obtained. The existence of electrical risks is determined based on the temporal and spatial characteristics of the winding impedance change value, the high-frequency transient energy value, and the spectral distortion value. Based on the electrical risk assessment results, the first candidate position is selected according to the distribution characteristics of the high-frequency transient energy value and the correlation characteristics of the winding impedance change value, the spectral distortion value and the abnormal magnetic flux distribution value. Obtain the insulation discharge intensity value of the insulating oil in the oil-immersed transformer; The second candidate position is determined based on the relevant characteristics of the spectral distortion value, the magnetic flux distribution anomaly value, the high-frequency transient energy value, and the insulation discharge intensity value. The electrical fault type is determined based on the temporal correlation characteristics of the first candidate position and the second candidate position within a preset time period; A risk warning is generated based on the electrical fault type.

2. The steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring as described in claim 1, characterized in that, Determining whether the state of the oil-immersed transformer is abnormal includes: Based on the temporal relationship between the winding impedance change value and the magnetic flux distribution anomaly value, a historical state vector is constructed, and the state mean is determined based on the historical state vector; according to the spatial distance between the historical state vector and the state mean, a historical vector distance set is determined. Based on the temporal relationship between the winding impedance change value and the magnetic flux distribution anomaly value obtained at the current moment, a current state vector is constructed, and the current state mean is determined based on the current state vector; the current vector distance is determined based on the spatial distance between the current state vector and the current state mean. The comparison distance is determined based on the spatial distance in the historical vector distance set corresponding to the preset level; Based on the comparison result between the current vector distance and the comparison distance, it is determined that the state of the oil-immersed transformer is abnormal.

3. The steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring as described in claim 2, characterized in that, The determination of whether an electrical risk exists includes: Based on the winding impedance change value, the high-frequency transient energy value, and the spectral distortion value at each moment within the next preset abnormal time period, time series sequences are constructed respectively to obtain impedance time series sequence, energy time series sequence, and distortion time series sequence. Based on the linear trends of the impedance time series, the energy time series, and the distortion time series, the impedance strength coefficient, energy strength coefficient, and distortion strength coefficient are quantitatively determined. The condition of simultaneous upward movement is determined based on the consistency of the signs of the impedance strength coefficient, energy strength coefficient, and distortion strength coefficient. Based on the aforementioned upward trend condition, impedance rate change, energy rate change, and distortion rate change are determined according to the respective rates of change of the impedance time series, the energy time series, and the distortion time series at adjacent moments within the preset abnormal duration. The existence of electrical risks is determined based on the impedance speed change, the energy speed change, and the distortion speed change.

4. The steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring as described in claim 3, characterized in that, Determining the existence of electrical risks based on the impedance speed change, the energy speed change, and the distortion speed change includes: Calculate the ratio of the number of times during which the impedance rate change, the energy rate change, and the distortion rate change are all positive within the preset abnormal duration to the total number of times during the preset abnormal duration, in order to obtain the synchronization rate of change. When the synchronous change rate is greater than a preset change threshold, an electrical risk is determined to exist.

5. The steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring as described in claim 4, characterized in that, The selection of the first candidate position includes: The cross-correlation degree is determined based on the cross-correlation relationship of the time series of the high-frequency transient energy values ​​at different preset spatial locations, and the spatial correlation is determined based on the threshold screening results of the cross-correlation degree. The preset spatial point corresponding to the maximum value among the high-voltage transient energy, low-voltage transient energy, and iron core transient energy determined based on the time-domain energy characteristics of the high-frequency transient energy values ​​at different preset spatial points is marked as the energy dominance point; The time sequence of recording the high-frequency transient energy values ​​of the energy-dominant point within the preset positioning time is the dominant energy sequence; The first candidate position is selected based on the spatial correlation and the correlation characteristics between the dominant energy sequence and the winding impedance change value, the spectral distortion value, and the magnetic flux distribution anomaly value.

6. The steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring as described in claim 5, characterized in that, The cross-correlation relationship of the time series of high-frequency transient energy values ​​based on different preset spatial locations is used to determine the cross-correlation degree, and spatial correlation is determined based on the threshold screening result of the cross-correlation degree, including: Based on the temporal characteristics of the high-frequency transient energy values ​​at the preset spatial locations, high-voltage energy sequences, low-voltage energy sequences, and core energy sequences are constructed respectively. The correlation between high and low pressure, high-pressure and low-pressure, and low-pressure are determined based on the correlation characteristics between the high-pressure energy sequence, the low-pressure energy sequence, and the core energy sequence, respectively. The spatial correlation is determined to be a strong spatial correlation based on the comparison results of the high-low pressure cross-correlation, the high-speed rail pressure cross-correlation, and the low-speed rail cross-correlation with the same threshold.

7. The steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring as described in claim 6, characterized in that, The step of filtering the first candidate position based on the spatial correlation and the correlation features between the dominant energy sequence and the winding impedance change value, the spectral distortion value, and the magnetic flux distribution anomaly value includes: Based on the time-series characteristics of the winding impedance change value, the spectral distortion value, and the magnetic flux distribution anomaly value, a winding resistance sequence, a harmonic distortion sequence, and a magnetic field strength sequence are constructed. The resistance oscillation correlation degree and the magnetic field oscillation correlation degree are determined according to the correlation degree between the dominant energy sequence and the winding resistance sequence and the magnetic field strength sequence, respectively. The wave resistance synchronization ratio is determined based on the cooperative characteristics of the changing trends of the winding resistance sequence and the harmonic distortion sequence within the preset positioning time. The harmonic distortion variance is determined based on the statistical fluctuation characteristics of the harmonic distortion sequence. Based on the strong spatial correlation, and according to the multi-threshold joint determination results of the resistance oscillation correlation, the magnetic field oscillation correlation, and the wave impedance synchronization ratio, the energy dominance point is determined as the first candidate position.

8. The steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring as described in claim 7, characterized in that, Determining the second candidate position includes: The magnetic resonance synergy index is determined based on the correlation between the local intensity sequence and the transient amplitude sequence constructed from the normalized temporal characteristics of the magnetic flux distribution anomaly and the high-frequency transient energy value, respectively. The rate of change of intensity is determined based on the degree of difference between the insulation discharge intensity value and the preset intensity threshold. The distortion rate is determined based on the degree of difference between the spectral distortion value and the preset harmonic threshold. The second candidate position is determined based on the threshold comparison results of the magnetic resonance synergy index, the intensity change rate, and the distortion change rate.

9. The steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring as described in claim 8, characterized in that, The second candidate position is determined based on a threshold comparison of the magnetic resonance symmetry index, the intensity change rate, and the distortion change rate, including: Based on the threshold screening results of the magnetic resonance synergy index, the risk type is determined to be strong magnetic circuit related type; Based on the threshold screening results of the magnetic resonance synergy index, the intensity change rate, and the distortion change rate, the second candidate position is determined to be an oil-paper composite insulation system.

10. The steady-state high-frequency electromagnetic spectrum analysis method based on power transmission and distribution facility condition monitoring as described in claim 9, characterized in that, The step of determining the electrical fault type based on the temporal correlation characteristics of the first candidate position and the second candidate position within a preset time period includes: The positional coordination ratio is determined statistically based on the coordination time characteristics of the first and second candidate positions. When the position coordination ratio is greater than a preset ratio threshold and the first candidate position is the energy dominance point determined based on the high voltage transient energy, the electrical fault type is determined to be a winding short circuit fault. When the position coordination ratio is greater than a preset ratio threshold and the first candidate position is the energy dominance point determined based on the low-voltage transient energy, the electrical fault type is determined to be a winding overload fault. When the positional coordination ratio is greater than a preset ratio threshold, and the first candidate position is the energy dominance point determined based on the transient energy of the iron core, the electrical fault type is determined to be an iron core flux leakage fault.

Images