A substation equipment operation data monitoring method and system

Abstract

The application relates to the technical field of electromagnetic measurement, and discloses a substation equipment operation data monitoring method and system, the method comprising the following steps: performing time-frequency feature analysis on an electromagnetic operation variable measurement data set of a substation equipment to obtain comprehensive electric variable feature parameters; constructing a time-space correlation graph according to timestamp information of the electromagnetic operation variable measurement data set and a spatial topological relation of the substation equipment; performing abnormality evaluation on the comprehensive electric variable feature parameters to obtain initial abnormal feature parameters; combining function load data of the substation equipment, and reconstructing the initial abnormal feature parameters into fault state monitoring information through electric-thermal coupling characteristic analysis; and fusing the time-space correlation graph and the fault state monitoring information into a state monitoring report; and the application can improve the monitoring efficiency of substation equipment operation data.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electromagnetic measurement, in particular to a substation equipment operation data monitoring method and system. BACKGROUND

[0002] As the core hub of power system power conversion and transmission, the real-time monitoring of the equipment operation state of the substation is a key link to ensure the safe and stable operation of the power network. With the continuous expansion of the scale of the power system and the improvement of the complexity of the equipment, the accuracy and timeliness of the substation equipment operation data monitoring are increasingly required, but the existing monitoring technology has obvious shortcomings in the data collection and integration stage. The traditional monitoring scheme mostly adopts a decentralized data collection mode, and each monitoring terminal respectively acquires single electrical parameters such as voltage and current, without realizing the synchronous collection of multi-dimensional electromagnetic variables such as voltage waveform, current waveform and dielectric loss factor, and lacking a unified time stamp marking and logical verification mechanism, resulting in problems such as time sequence confusion and data defects in the collected data set, which cannot form a complete and strongly correlated electromagnetic operation variable data set, and poses hidden dangers for subsequent monitoring analysis.

[0003] The technical defects of the existing monitoring technology in the data analysis and fault judgment link further restrict the improvement of the monitoring efficiency. In the feature parameter extraction stage, the traditional scheme does not carry out systematic time-frequency domain joint analysis, nor does it establish a scientific feature saliency evaluation system, making it difficult to eliminate redundant feature parameters, so that the extracted electrical variable feature dimension is redundant and the core information is not prominent, which cannot accurately reflect the key state of the equipment operation; at the same time, the existing technology does not construct a time-space correlation model combined with the spatial topological relationship of the equipment, cannot realize the linkage monitoring of the operation state between the equipment, and the abnormal evaluation only relies on a single dielectric loss factor parameter, without combining with the equipment function load data for electrical-thermal coupling characteristic analysis, which cannot accurately divide the abnormal level and is also difficult to locate the root cause of the fault, ultimately resulting in insufficient practicality and accuracy of the monitoring report, which cannot meet the fine and efficient monitoring needs of the substation equipment, so how to improve the monitoring efficiency of the substation equipment operation data has become a problem to be solved. SUMMARY

[0004] The present application provides a substation equipment operation data monitoring method and system to solve the problems raised in the background art.

[0005] To achieve the above purpose, the present application provides a substation equipment operation data monitoring method, which comprises:

[0006] S1, integrating the voltage waveform measurement data, the current waveform measurement data and the dielectric loss factor measurement data of the substation equipment into the electromagnetic operation variable measurement data set of the substation equipment;

[0007] S2, time-frequency feature analysis is performed on the electromagnetic operation variable measurement data set to obtain comprehensive electrical variable feature parameters of the substation equipment;

[0008] S3, a space-time correlation graph of the substation equipment is constructed according to timestamp information of the electromagnetic operation variable measurement data set and a spatial topological relationship of the substation equipment;

[0009] S4, abnormality evaluation is performed on the comprehensive electrical variable feature parameters based on the dielectric loss factor measurement data to obtain initial abnormal feature parameters of the substation equipment;

[0010] S5, the initial abnormal feature parameters are reconstructed into fault state monitoring information of the substation equipment through electrical-thermal coupling characteristic analysis in combination with functional load data of the substation equipment;

[0011] S6, the space-time correlation graph and the fault state monitoring information are associated and fused into state monitoring report data of the substation equipment.

[0012] In a preferred embodiment, the integration of voltage waveform measurement data, current waveform measurement data and dielectric loss factor measurement data of the substation equipment into the electromagnetic operation variable measurement data set of the substation equipment comprises:

[0013] Synchronously measuring voltage waveform measurement data, current waveform measurement data and dielectric loss factor measurement data of the substation equipment;

[0014] Adding timestamps to the voltage waveform measurement data, the current waveform measurement data and the dielectric loss factor measurement data to obtain a time-stamped electromagnetic measurement data set of the substation equipment;

[0015] Performing logical verification on the time-stamped electromagnetic measurement data set to obtain the electromagnetic operation variable measurement data set of the substation equipment.

[0016] In a preferred embodiment, the time-frequency feature analysis on the electromagnetic operation variable measurement data set to obtain the comprehensive electrical variable feature parameters of the substation equipment comprises:

[0017] Performing time-frequency domain joint analysis on voltage waveform measurement data and current waveform measurement data in the electromagnetic operation variable measurement data set to obtain voltage time-frequency feature matrices and current time-frequency feature matrices of the substation equipment;

[0018] Based on the voltage time-frequency feature matrices and the current time-frequency feature matrices, phase difference between voltage and current in the substation equipment is evaluated to obtain phase difference candidate feature parameters of the substation equipment;

[0019] The frequency domain variation characteristics of the dielectric loss factor measurement data in the electromagnetic operating variable measurement dataset are used as candidate characteristic parameters of dielectric loss of the substation equipment.

[0020] Redundant feature parameters are removed from the candidate feature parameters of phase difference and candidate feature parameters of dielectric loss to obtain the core candidate feature parameters of the substation equipment.

[0021] The core candidate feature parameters are dimensionality reduced and fused to obtain the comprehensive electrical variable feature parameters of the substation equipment.

[0022] In a preferred embodiment, the step of removing redundant feature parameters from the candidate feature parameters of the phase difference and the candidate feature parameters of the dielectric loss to obtain the core candidate feature parameters of the substation equipment includes:

[0023] The phase difference candidate feature parameters and the dielectric loss candidate feature parameters are integrated into a candidate feature parameter set for the substation equipment;

[0024] Obtain the numerical distribution range of the candidate feature parameter set under the historical normal operation status and typical fault modes of the substation equipment, as well as the current equipment operation status;

[0025] Based on the numerical distribution range and the current device operating status, the feature saliency intensity of the candidate feature parameter set is calculated, wherein the formula for calculating the feature saliency intensity is as follows:

[0026] ;

[0027] In the formula, Indicates the first in the candidate feature parameter set The feature significance strength of each candidate feature parameter Indicates the first in the candidate feature parameter set The current monitoring value of each candidate feature parameter. Indicates the first in the candidate feature parameter set The average value of each candidate feature parameter under the historical normal operating conditions. Indicates the first in the candidate feature parameter set The standard deviation of each candidate feature parameter under the historical normal operating conditions This represents a preset weighting coefficient used to characterize the deviation of the current monitored value from the normal historical range. Indicates the first in the candidate feature parameter set The mean of the candidate characteristic parameters under the typical failure mode. Indicates the first in the candidate feature parameter set The standard deviation of each candidate characteristic parameter under the typical failure mode a function for finding the maximum value in a set of numerical values, a preset weight coefficient for characterizing that the current monitoring value approaches the failure interval;

[0028] determining a candidate feature parameter in the candidate feature parameter set as a redundant feature parameter if the feature saliency strength of the candidate feature parameter is lower than a preset saliency threshold;

[0029] eliminating the redundant feature parameter from the candidate feature parameter set to obtain a core candidate feature parameter of the substation equipment.

[0030] In a preferred embodiment, constructing the space-time correlation graph of the substation equipment according to the timestamp information of the electromagnetic operation variable measurement data set and the spatial topological relationship of the substation equipment comprises:

[0031] constructing a graph structure basis of the substation equipment according to the spatial topological relationship of the substation equipment, taking the substation equipment as nodes and the connection relationship between the devices in the substation equipment as edges;

[0032] mapping the electromagnetic operation variable measurement data to the nodes in the graph structure basis based on the timestamp information in the electromagnetic operation variable measurement data to obtain a time correlation graph of the substation equipment;

[0033] inputting the weight attribute in the spatial topological relationship to the edges in the time correlation graph to obtain the space-time correlation graph of the substation equipment.

[0034] In a preferred embodiment, the abnormality evaluation of the comprehensive electric variable feature parameter based on the dielectric loss factor measurement data obtains an initial abnormal feature parameter of the substation equipment, which comprises:

[0035] determining a dielectric loss factor change reference of the dielectric loss factor measurement data according to the historical time series data of the substation equipment;

[0036] performing correlation analysis on the comprehensive electric variable feature parameter and the dielectric loss factor change reference to construct an abnormality judgment rule of the comprehensive electric variable feature parameter;

[0037] differentially comparing the deviation degree between the dielectric loss factor measurement data and the dielectric loss factor change reference to obtain a deviation state of the dielectric loss factor measurement data;

[0038] jointly identifying the abnormal state in the abnormality judgment rule and the deviation state to obtain an abnormal preliminary conclusion of the substation equipment;

[0039] According to the abnormal preliminary judgment conclusion, an initial abnormal level of the comprehensive electric variable characteristic parameter is divided;

[0040] The abnormal preliminary judgment conclusion and the initial abnormal level are parameterized and packaged to obtain an initial abnormal characteristic parameter of the substation equipment.

[0041] In a preferred embodiment, the correlation analysis of the comprehensive electric variable characteristic parameter and the dielectric loss factor change reference is performed to construct an abnormal judgment rule of the comprehensive electric variable characteristic parameter, including:

[0042] According to a reference threshold interval and a trend change characteristic in the dielectric loss factor change reference, a dielectric loss reference correlation element of the substation equipment is determined;

[0043] According to the parameter types of voltage, current and phase difference, the comprehensive electric variable characteristic parameter is classified to obtain a classified electric variable characteristic subset of the substation equipment;

[0044] According to a correlation relationship between the classified electric variable characteristic subset and the dielectric loss reference correlation element, a parameter reference correlation mapping table of the substation equipment is constructed;

[0045] Based on the parameter reference correlation mapping table, an abnormal trigger condition of the classified electric variable characteristic subset is determined;

[0046] The abnormal trigger condition is integrated to obtain an abnormal judgment rule of the comprehensive electric variable characteristic parameter.

[0047] In a preferred embodiment, the initial abnormal characteristic parameter is reconstructed into fault state monitoring information of the substation equipment by electric-thermal coupling characteristic analysis in combination with functional load data of the substation equipment, including:

[0048] Current load type, load duration and load fluctuation amplitude of the substation equipment are collected to obtain a load characteristic data set of the substation equipment;

[0049] An abnormal parameter type and an abnormal degree in the initial abnormal characteristic parameter are taken as an abnormal core element of the substation equipment;

[0050] The load characteristic data set and the abnormal core element are subjected to electric-thermal coupling characteristic analysis to establish a correlation and correspondence relationship between the load characteristic data set and the abnormal core element;

[0051] Based on the correlation and correspondence relationship, a device fault inducement corresponding to the initial abnormal characteristic parameter is determined to obtain a fault inducement determination result of the substation equipment;

[0052] The initial abnormal feature parameter, the fault cause judgment result and the load feature data set are combined to obtain the fault state monitoring information of the substation equipment.

[0053] In a preferred embodiment, the association and fusion of the space-time correlation graph and the fault state monitoring information into the state monitoring report of the substation equipment comprises:

[0054] The device node identifier, node electrical parameter data and inter-device topology correlation weight in the space-time correlation graph are analyzed to obtain a set of space-time monitoring elements of the substation equipment.

[0055] The fault state monitoring information is structurally split to obtain fault core information of the substation equipment.

[0056] The set of space-time monitoring elements and the fault core information are bound to obtain a set of space-time fault correlation data of the substation equipment.

[0057] The set of space-time fault correlation data is time-space sorted according to the device area and fault level of the substation equipment to obtain a sequence of space-time fault correlation data of the substation equipment.

[0058] The sequence of space-time fault correlation data is deconstructed and compiled to form the state monitoring report of the substation equipment.

[0059] To solve the above problems, the application further provides a substation equipment operation data monitoring system, which comprises:

[0060] A data acquisition module is configured to integrate voltage waveform measurement data, current waveform measurement data and dielectric loss factor measurement data of a substation equipment into an electromagnetic operation variable measurement data set of the substation equipment.

[0061] A time-frequency domain analysis module is configured to perform time-frequency feature analysis on the electromagnetic operation variable measurement data set to obtain comprehensive electrical variable feature parameters of the substation equipment.

[0062] A space correlation graph construction module is configured to construct a space-time correlation graph of the substation equipment according to timestamp information of the electromagnetic operation variable measurement data set and a spatial topology relationship of the substation equipment.

[0063] An abnormality evaluation module is configured to perform abnormality evaluation on the comprehensive electrical variable feature parameters based on the dielectric loss factor measurement data to obtain initial abnormal feature parameters of the substation equipment.

[0064] An electro-thermal coupling analysis module is configured to combine function load data of the substation equipment, and reconstruct the initial abnormal characteristic parameter into fault state monitoring information of the substation equipment through electro-thermal coupling characteristic analysis.

[0065] A monitoring report generation module is configured to associate and fuse the space-time correlation graph with the fault state monitoring information into a state monitoring report of the substation equipment.

[0066] Compared with the prior art, the present application has the following beneficial effects:

[0067] 1. The present application forms an electromagnetic operation variable data set by integrating voltage waveform, current waveform and dielectric loss factor multidimensional data, extracts characteristic parameters through joint analysis in time and frequency domains, and removes redundant parameters relying on a characteristic saliency strength formula, thereby guaranteeing the accuracy of core characteristics; meanwhile, a space-time correlation graph is constructed by combining time stamp and equipment space topology, thereby realizing space-time linkage monitoring of equipment operation state, breaking the information island problem of traditional decentralized monitoring, and making monitoring data have stronger correlation and completeness.

[0068] 2. The present application establishes abnormal judgment rules based on dielectric loss factor, completes abnormal preliminary judgment and grade division of comprehensive electric variable characteristics, and then carries out electro-thermal coupling characteristic analysis combined with equipment load data, thereby accurately locating fault causes; finally, the space-time correlation graph and fault information are fused to generate a structured monitoring report according to equipment area and fault grade, thereby providing monitoring personnel with a monitoring conclusion with space-time dimension and fault root, and greatly improving the pertinence of fault disposal and the overall efficiency of monitoring work. BRIEF DESCRIPTION OF DRAWINGS

[0069] Figure 1 A flowchart of a substation equipment operation data monitoring method provided by an embodiment of the present application is shown;

[0070] Figure 2 A functional module diagram of a substation equipment operation data monitoring system provided by an embodiment of the present application is shown;

[0071] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0072] It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0073] The embodiment of the present application provides a kind of substation equipment operation data monitoring method.The execution subject of the kind of substation equipment operation data monitoring method includes but is not limited to server, terminal and at least one of the electronic device that can be configured to execute the method provided in the application of the present application, etc.It is said in other words, the kind of substation equipment operation data monitoring method can be executed by software or hardware installed in terminal equipment or server equipment.The server includes but is not limited to: single server, server cluster, cloud server or cloud server cluster, etc.The server can be independent server, can also be cloud server that provides cloud service, cloud database, cloud computing, cloud function, cloud storage, network service, cloud communication, middleware service, domain name service, security service, content distribution network (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platform.

[0074] Referring to Figure 1 As shown in the figure, a flowchart of a kind of substation equipment operation data monitoring method provided by an embodiment of the present application.In this embodiment, the kind of substation equipment operation data monitoring method includes:

[0075] S1, the voltage waveform measurement data, current waveform measurement data and dielectric loss factor measurement data of substation equipment are integrated into the electromagnetic operating variable measurement data set of the substation equipment;

[0076] In the embodiment of the present application, the voltage waveform measurement data, current waveform measurement data and dielectric loss factor measurement data of substation equipment are integrated into the electromagnetic operating variable measurement data set of the substation equipment, including:

[0077] Synchronously measure the voltage waveform measurement data, current waveform measurement data and dielectric loss factor measurement data of substation equipment;

[0078] Time stamp is added to the voltage waveform measurement data, current waveform measurement data and dielectric loss factor measurement data, to obtain the electromagnetic measurement data set with time mark of the substation equipment;

[0079] Logical check is carried out on the electromagnetic measurement data set with time mark, to obtain the electromagnetic operating variable measurement data set of the substation equipment.

[0080] In the measurement area of the substation equipment, a unified Beidou timing module is first deployed as a time reference source for all devices participating in data collection. The module can receive high-precision time signals from Beidou satellites and synchronously output standard time bases to each collection device. At the same time, the voltage collection module is connected to the secondary side of the voltage transformer of the substation equipment through a dedicated wiring, the current collection module is connected to the secondary side of the current transformer of the equipment, and the dielectric loss factor collection unit establishes a data transmission link with the voltage collection module and the current collection module. Hardware deployment and link debugging before collection are completed, and then all collection devices are started. The voltage collection module continuously samples the voltage signals transmitted by the voltage transformer at a fixed collection interval, converts the sampled voltage signals into voltage waveform measurement data, and the current collection module synchronously samples the current signals transmitted by the current transformer at the same collection interval to form current waveform measurement data. The dielectric loss factor collection unit receives voltage waveform measurement data and current waveform measurement data at the same time in real time, and obtains dielectric loss factor measurement data at the corresponding time by analyzing the phase and amplitude relationship of the two types of data in real time. During the entire collection process, all devices rely on the unified time base of the Beidou timing module to ensure that the start time and collection interval are completely consistent, thereby realizing the synchronous acquisition of voltage waveform measurement data, current waveform measurement data and dielectric loss factor measurement data of the substation equipment.

[0081] After the synchronous collection work is completed, the accurate time information of each collection time recorded by the Beidou timing module is retrieved. For each piece of collected voltage waveform measurement data, the time information corresponding to the actual collection time is associated with each waveform sampling point data carrier of the data. For each piece of current waveform measurement data, the time information corresponding to the collection time is associated with each waveform sampling point data carrier in the same way. For each piece of dielectric loss factor measurement data, the time information corresponding to the calculation generation time is directly associated with the carrier of the data. After completing the timestamp addition operation of all data, the voltage waveform measurement data with added timestamp, the current waveform measurement data with added timestamp and the dielectric loss factor measurement data with added timestamp are classified and integrated according to the collection time to form a time-stamped electromagnetic measurement data set of the substation equipment containing three types of time-stamped data.

[0082] A dedicated logical verification system is first built, which includes a data integrity verification link, a time sequence consistency verification link and a data rationality verification link. After entering the verification process, a single set of to-be-verified data is extracted from the time-stamped electromagnetic measurement data set. In the data integrity verification link, it is checked whether the voltage waveform measurement data, the current waveform measurement data and the dielectric loss factor measurement data are contained in the data set, and it is confirmed that there is no data missing section and no carrier damage in the carrier of each type of data, so as to ensure the integrity of the basic form of the three types of data. Then, in the time sequence consistency verification link, the time stamps of the three types of data in the data set are compared, and it is confirmed that the time difference of the time stamps of the voltage waveform measurement data, the current waveform measurement data and the dielectric loss factor measurement data at the same collection time meets the preset standard of synchronous measurement. If the time stamp deviation exceeds the standard, it is determined that the time sequence is abnormal data and is marked and removed. Then, in the data rationality verification link, the historical operation data and the factory technical characteristic data of the substation equipment are called, and the voltage waveform amplitude and form and the current waveform amplitude and form in the to-be-verified data are compared with the waveform characteristics in the normal operation state of the equipment. At the same time, the value of the dielectric loss factor measurement data is compared with the normal range of the dielectric loss factor in the normal operation of the equipment. If the voltage waveform has an abrupt amplitude change without basis, the current waveform has a form distortion without fault state, and the dielectric loss factor has an abrupt value increase without inducement, it is determined that the data is rationality abnormal data and is marked and removed. After three rounds of verification of all the data in the time-stamped electromagnetic measurement data set, all the data passing the verification are re-regularized according to the collection time, and finally the electromagnetic operation variable measurement data set of the substation equipment is obtained.

[0083] S2, time-frequency feature analysis is performed on the electromagnetic operation variable measurement data set to obtain comprehensive electric variable feature parameters of the substation equipment.

[0084] In the embodiment of the application, the time-frequency feature analysis on the electromagnetic operation variable measurement data set to obtain comprehensive electric variable feature parameters of the substation equipment includes:

[0085] The voltage waveform measurement data and the current waveform measurement data in the electromagnetic operation variable measurement data set are subjected to joint analysis in the time-frequency domain to obtain a voltage time-frequency feature matrix and a current time-frequency feature matrix of the substation equipment.

[0086] Based on the voltage time-frequency feature matrix and the current time-frequency feature matrix, the phase difference between the voltage and the current in the substation equipment is evaluated to obtain a phase difference candidate feature parameter of the substation equipment.

[0087] The frequency domain variation feature of the dielectric loss factor measurement data in the electromagnetic operation variable measurement data set is taken as a dielectric loss candidate feature parameter of the substation equipment.

[0088] redundant feature parameters in the phase difference candidate feature parameters and the dielectric loss candidate feature parameters to obtain core candidate feature parameters of the substation equipment;

[0089] perform dimension reduction fusion on the core candidate feature parameters to obtain comprehensive electrical variable feature parameters of the substation equipment.

[0090] The removing redundant feature parameters in the phase difference candidate feature parameters and the dielectric loss candidate feature parameters to obtain core candidate feature parameters of the substation equipment comprises:

[0091] integrating the phase difference candidate feature parameters and the dielectric loss candidate feature parameters into a candidate feature parameter set of the substation equipment;

[0092] obtaining numerical distribution ranges of the candidate feature parameter set under a historical normal operation state and a typical fault mode of the substation equipment and a current equipment operation state;

[0093] calculating feature saliency strengths of the candidate feature parameter set according to the numerical distribution ranges and the current equipment operation state, wherein a calculation formula of the feature saliency strengths is as follows:

[0094] ;

[0095] In the formula, S i represents a feature saliency strength of an i-th candidate feature parameter in the candidate feature parameter set, S i represents a current monitoring value of an i-th candidate feature parameter in the candidate feature parameter set, S i represents a mean value of an i-th candidate feature parameter in the candidate feature parameter set under the historical normal operation state, S i represents a standard deviation of an i-th candidate feature parameter in the candidate feature parameter set under the historical normal operation state, S i represents a preset weight coefficient for representing that the current monitoring value deviates from a normal historical interval, S i represents a mean value of an i-th candidate feature parameter in the candidate feature parameter set under the typical fault mode, S i represents a standard deviation of an i-th candidate feature parameter in the candidate feature parameter set under the typical fault mode, represents a function for obtaining a maximum value in a group of numerical values, represents a preset weight coefficient for representing that the current monitoring value approaches a fault interval; ​​​​​​​

[0096] determining a candidate feature parameter with a feature saliency intensity lower than a preset saliency threshold in the candidate feature parameter set as a redundant feature parameter;

[0097] eliminating the redundant feature parameter from the candidate feature parameter set to obtain a core candidate feature parameter of the substation equipment.

[0098] All voltage waveform measurement data and current waveform measurement data in the electromagnetic operation variable measurement dataset are retrieved, time domain feature extraction is first carried out on the voltage waveform measurement data, the voltage waveform data is divided into continuous and non-overlapping time windows according to a fixed time length, in each time window, time domain features such as the amplitude peak value, amplitude mean value, waveform rising edge slope mean value and waveform falling edge slope mean value of the voltage waveform in the window are counted, then frequency domain feature extraction is carried out on the voltage waveform data in each time window, by converting the time domain waveform signal into the distribution form of different frequency components, frequency domain features such as the energy proportion of each frequency component, dominant frequency value and distribution concentration degree of the frequency component are counted, then the time domain features and frequency domain features corresponding to each time window are integrated into a matrix row vector according to a preset arrangement order, the row vectors corresponding to all time windows are arranged in time sequence, forming a voltage time-frequency feature matrix of the substation equipment, the same operation process is used to extract and integrate the time domain and frequency domain features of the current waveform measurement data, finally a current time-frequency feature matrix of the substation equipment is obtained.

[0099] First, the voltage time-frequency feature vector and the current time-frequency feature vector corresponding to the same time window are extracted from the voltage time-frequency feature matrix and the current time-frequency feature matrix, the core frequency component of the voltage waveform in the time window is locked according to the dominant frequency value in the voltage time-frequency feature vector, then the feature information corresponding to the same core frequency component is matched from the current time-frequency feature vector, by comparing the phase correlation characteristics of voltage and current under the core frequency component, the phase shift degree of the voltage waveform and the current waveform in the time window is determined, then the phase shift degree of voltage and current in all time windows is evaluated one by one in the above manner, at the same time, the core frequency phase shift value corresponding to all time windows, the dispersion degree of phase shift of each frequency component and the change trend of phase shift with time are integrated, forming a phase difference candidate feature parameter of the substation equipment.

[0100] The all medium loss factor measurement data in the electromagnetic operation variable measurement data set is called. The medium loss factor measurement data is divided into each time window according to the same time window division rule as the voltage and current time frequency analysis. In each time window, the numerical value change of the medium loss factor under different frequency components is analyzed in combination with the voltage and current frequency domain characteristics corresponding to the window. The medium loss factor value corresponding to each frequency component, the slope of the medium loss factor with frequency, the numerical fluctuation amplitude of the medium loss factor in the window, and the average value of the medium loss factor in the window are counted. The medium loss factor frequency domain change characteristics corresponding to all time windows are collected in time sequence to form the medium loss candidate characteristic parameters of the substation equipment.

[0101] The generated phase difference candidate characteristic parameters and medium loss candidate characteristic parameters of the substation equipment are called. All feature items of the two types of characteristic parameters are normalized in a unified format. A phase feature exclusive identification is added to each feature item in the phase difference candidate characteristic parameters, and a loss feature exclusive identification is added to each feature item in the medium loss candidate characteristic parameters. Then, the feature items are classified according to the equipment electromagnetic state dimensions reflected by the feature items. The feature items reflecting the phase static offset characteristics are classified into the phase static dimension, the feature items reflecting the phase dynamic change characteristics are classified into the phase dynamic dimension, the feature items reflecting the medium loss basic numerical value characteristics are classified into the loss numerical value dimension, and the feature items reflecting the medium loss frequency response characteristics are classified into the loss frequency response dimension. Subsequently, all classified feature items are arranged in dimension order, and a unified feature index directory is established for the entire feature set. The index directory includes the identification, dimension, corresponding data source and feature meaning of each feature item. After the above operations, the phase difference candidate characteristic parameters and the medium loss candidate characteristic parameters that are normalized, classified and indexed are collected as a whole to form the candidate characteristic parameter set of the substation equipment.

[0102] The historical operation database and the fault case database of the first access substation equipment are accessed, all monitoring records of the equipment in the normal operation state in the past five years are called from the historical operation database, the historical values of all feature items corresponding to the candidate feature parameter set are extracted, the monitoring records corresponding to various typical fault modes of the equipment are called from the fault case database, including insulation aging fault, partial discharge fault, winding deformation fault, etc., the fault values of all feature items corresponding to the candidate feature parameter set are also extracted, for each feature item in the candidate feature parameter set, the value fluctuation interval, the value mean and the value dispersion degree in the historical normal operation state are respectively counted to form the value distribution range of the feature item in the normal state, the value fluctuation interval, the value mean and the deviation amplitude from the normal state value in various typical fault modes are respectively counted to form the value distribution range of the feature item in each typical fault mode, the current operation voltage level, load rate, environmental temperature and humidity and other basic operation parameters of the equipment are called through the real-time monitoring system of the first access substation equipment, combined with the real-time inspection records of the field operation and maintenance personnel, including information such as whether the equipment has abnormal sound and leakage marks, the current equipment operation state of the substation equipment is comprehensively integrated.

[0103] The feature significance intensity determination logic is first established, which takes the distinguishing ability of the feature item to the equipment state as the core determination basis, for each feature item in the candidate feature parameter set, the current value of the feature item is compared with the value distribution range in the historical normal operation state, the degree of deviation of the current value from the normal mean is calculated, the current value is compared with the value distribution range in various typical fault modes, the state interval to which the current value belongs is determined, then the difference amplitude of the feature item between the normal state and various typical fault modes is evaluated, the larger the difference amplitude is, the higher the distinguishing degree of the feature item to the state is, at the same time, the numerical distinguishing degree of the feature item between different typical fault modes is evaluated, the higher the distinguishing degree is, the stronger the recognition degree of the feature item to the fault type is, finally, the deviation degree of the current value, the difference amplitude between the states and the fault type recognition degree are comprehensively integrated to determine the feature significance intensity corresponding to the feature item, after the intensity calculation of all feature items is completed, the feature significance intensity set corresponding to the candidate feature parameter set is formed.

[0104] The current monitoring value is the current monitoring value of the candidate characteristic parameter in the candidate characteristic parameter set, which is obtained by accessing the real-time monitoring system of the substation equipment, calling the real-time monitoring data of the candidate characteristic parameter under the current operating state of the equipment, and checking the data in combination with the real-time inspection records of the field operation and maintenance personnel. The mean value under the historical normal operating state is obtained by calling all historical monitoring records of the corresponding candidate characteristic parameter under the normal operating state of the equipment in the past five years from the historical operation database of the substation equipment, and calculating the average value of the numerical values of the records. The standard deviation under the historical normal operating state is obtained by calculating the dispersion of the values of the candidate characteristic parameter under the normal operating state based on all historical monitoring records in the historical operation database. The mean value under the typical fault mode is obtained by calling all monitoring records of the corresponding candidate characteristic parameter under various typical fault modes of the equipment from the fault case database of the substation equipment, and calculating the average value of the numerical values of the records. The standard deviation under the typical fault mode is obtained by calculating the dispersion of the values of the candidate characteristic parameter under various typical fault modes based on all monitoring records in the fault case database. The preset weight coefficient representing the deviation of the current monitoring value from the normal historical interval is a fixed coefficient preset based on the historical characteristic analysis data of the same type of substation equipment and the state recognition requirement. The preset weight coefficient representing the approach of the current monitoring value to the fault interval is a fixed coefficient preset based on the fault characteristic analysis data of the same type of substation equipment and the fault recognition requirement.

[0105] The formula is used to calculate the characteristic significance strength of the candidate characteristic parameter in the candidate characteristic parameter set. Specifically, it obtains the strength by integrating two parts, the first part is the degree of deviation of the current monitoring value from the mean value under the historical normal operating state, and the second part is the degree of approach of the current monitoring value to the mean value under the typical fault mode. After assigning the corresponding preset weight to the two parts, the final result can reflect the ability of the candidate characteristic parameter to distinguish the historical normal operating state of the substation equipment from the typical fault mode. The stronger the distinguishing ability, the higher the corresponding characteristic significance strength value.

[0106] When the current monitoring value of the candidate feature parameter deviates from the mean value under the normal historical operating state to a greater extent, the value of the corresponding part in the formula will increase, thereby increasing the value of the feature saliency strength. When the current monitoring value of the candidate feature parameter approaches the mean value under the typical fault mode, the value of the corresponding part in the formula will increase, thereby increasing the value of the feature saliency strength. When the preset weight coefficient proportion representing the deviation of the current monitoring value from the normal historical interval is higher, the influence of the deviation of the current monitoring value from the normal mean value on the feature saliency strength will be greater. When the preset weight coefficient proportion representing the approach of the current monitoring value to the fault interval is higher, the influence of the approach of the current monitoring value to the fault mean value on the feature saliency strength will be greater.

[0107] First, a preset saliency threshold is determined, which is a fixed decision criterion preset based on historical feature analysis data of the same type of substation equipment and state recognition requirements. The value of the threshold corresponds to the minimum feature discrimination ability that can effectively distinguish between normal and fault states of the equipment. Then, the feature saliency strength corresponding to each feature item in the candidate feature parameter set is retrieved one by one, and the strength value is directly compared with the preset saliency threshold. If the feature saliency strength of a certain feature item is lower than the preset saliency threshold, it means that the feature item cannot effectively distinguish between the normal operating state and the typical fault mode of the equipment, and does not have effective support value for equipment state evaluation. At this time, the feature item is determined as a redundant feature parameter and is marked with a special redundant identifier. After completing the threshold comparison and identification of all feature items, a redundant feature parameter list corresponding to the candidate feature parameter set is formed.

[0108] First, all feature items are retrieved from the candidate feature parameter set, and the redundant feature parameter list generated is used to filter out all feature items that are not marked as redundant feature parameters. Then, a secondary format check is performed on these unmarked feature items to confirm that the identification, dimension, and data format of each feature item are complete and standardized. Subsequently, the feature items are classified and arranged again according to the equipment electromagnetic state dimension to which they belong, maintaining the classification logic of the phase static dimension, phase dynamic dimension, loss value dimension, and loss frequency response dimension. At the same time, the feature index directory is updated, and the related entries corresponding to the redundant feature parameters in the directory are removed. The latest dimension attribution information of the remaining feature items is supplemented and retained. Finally, the remaining feature items that have been arranged and updated are collected as a whole to form the core candidate feature parameters of the substation equipment.

[0109] First, all feature items in the core candidate feature parameters are classified by feature hierarchy. Feature items reflecting the absolute value of phase offset are classified into the phase absolute value level, feature items reflecting the trend of phase offset change are classified into the phase trend level, feature items reflecting the magnitude of dielectric loss are classified into the loss value level, and feature items reflecting the amplitude of dielectric loss change are classified into the loss fluctuation level. Then, for each feature level, the information contribution weight of all feature items in that level is calculated, and the feature item with the highest information contribution weight is selected as the representative feature of that level. Subsequently, the representative features of each level are integrated according to the logical order of equipment electromagnetic state assessment, and the scattered multi-dimensional feature information is integrated into a unified feature set that can comprehensively reflect the electromagnetic operating state of substation equipment. Finally, the comprehensive electrical variable feature parameters of substation equipment are obtained.

[0110] The beneficial effects are as follows: By conducting joint time-frequency domain analysis on voltage and current waveform measurement data in the electromagnetic operating variable measurement dataset, the original waveform data can be transformed into voltage and current time-frequency feature matrices that possess both time-domain variation characteristics and frequency-domain distribution features, providing multi-dimensional and accurate feature support for subsequent phase difference assessment. Evaluating the phase difference between voltage and current based on the voltage and current time-frequency feature matrices and obtaining candidate phase difference feature parameters can accurately capture the phase correlation characteristics of equipment voltage and current, laying a data foundation for phase dimension analysis of the equipment's electromagnetic state. Using the frequency domain variation characteristics of dielectric loss factor measurement data as candidate dielectric loss feature parameters can effectively uncover the variation law of dielectric loss with frequency, supplementing the core features related to the equipment's insulation characteristics. By eliminating redundant feature parameters in the phase difference and dielectric loss candidate feature parameters, invalid and repetitive feature information can be filtered out, ensuring that the core candidate feature parameters have high information density and strong relevance, and avoiding interference from redundant features in subsequent analysis. By performing dimensionality reduction and fusion on the core candidate feature parameters to obtain comprehensive electrical variable feature parameters, it is possible to organically integrate multi-dimensional features and form a unified feature set that can comprehensively reflect the electromagnetic operating status of equipment. This provides a highly integrated and accurate feature basis for substation equipment operating status assessment, potential fault early warning and other tasks.

[0111] S3. Based on the timestamp information of the electromagnetic operating variable measurement dataset and the spatial topology relationship of the substation equipment, construct the spatiotemporal correlation diagram of the substation equipment;

[0112] In this embodiment of the invention, constructing a spatiotemporal correlation diagram of the substation equipment based on the timestamp information of the electromagnetic operating variable measurement dataset and the spatial topology relationship of the substation equipment includes:

[0113] Based on the spatial topology of the substation equipment, a graph structure foundation for the substation equipment is constructed, with the substation equipment as nodes and the connection relationships between the equipment as edges.

[0114] Based on the timestamp information in the electromagnetic operating variable measurement data, the electromagnetic operating variable measurement data is mapped to the nodes in the graph structure foundation to obtain the time association graph of the substation equipment;

[0115] By inputting the weight attributes in the spatial topology into the edges of the temporal association graph, the spatiotemporal association graph of the substation equipment is obtained.

[0116] First, retrieve the electrical wiring diagrams, equipment physical layout ledgers, and cable connection lists of the substation. These documents must cover the deployment areas, equipment models, and specific interconnection paths of all substation equipment. Identify each independently operating substation device, including transformers, circuit breakers, disconnectors, busbars, etc., and designate each independent substation device as a node. Next, for each identified node, trace its actual connection relationships with the corresponding equipment of other nodes. For example, a busbar node and a circuit breaker node are directly connected by cables, and a circuit breaker node and a transformer node are directly connected by busbars. For each such direct connection relationship identified, establish an edge between the corresponding two nodes, ensuring that each edge accurately corresponds to a set of actual connections between devices, without omissions or incorrect associations. After identifying all nodes and establishing all corresponding edges, the structural set composed of these nodes and edges forms the graphical structural basis of the substation equipment.

[0117] First, the electromagnetic operating variable measurement dataset of the substation equipment is retrieved. This dataset contains voltage waveform measurement data, current waveform measurement data, and dielectric loss factor measurement data that have undergone logical verification. It also contains the timestamp information corresponding to each data item. The timestamp information corresponding to each data item and the specific substation equipment corresponding to that data item are extracted. Then, the node corresponding to the specific substation equipment in the graph structure is located. The extracted timestamp information and the electromagnetic operating variable measurement data corresponding to that timestamp are directly associated and bound to that node. For the same node, if there are multiple electromagnetic operating variable measurement data corresponding to different timestamps, these data are associated to the node in the order of the timestamps, ensuring that the measurement data corresponding to each timestamp can be accurately matched to the corresponding node. After completing the association operation of all data items with the corresponding nodes, the graph structure that originally only contained equipment nodes and connecting edges is given the electromagnetic operating variable measurement data information of each node at different times, thus obtaining the time association graph of the substation equipment.

[0118] First, the weight attributes in the spatial topology of the substation equipment are retrieved. These attributes are determined based on the spatial connection characteristics between the equipment, specifically including the rated current carrying capacity of the connecting cables between the equipment, the actual electrical distance between the equipment, the inherent impedance level of the connection circuit, and the mechanical bearing capacity of the connecting components. This attribute information needs to be extracted from the electrical parameter ledger, cable technical specifications, and equipment installation records of the substation. Then, each edge in the time correlation graph is located. Each edge corresponds to the connection relationship between two substation equipment nodes. The weight attribute information corresponding to this connection relationship is input and bound to this edge one by one. For multiple weight attributes corresponding to each edge, all attribute information is integrated and uniformly associated with this edge to ensure that each edge fully carries the corresponding spatial topology weight information. After completing the binding operation of all edges and corresponding weight attributes, the time correlation graph simultaneously possesses topology weight information in the spatial dimension and electromagnetic operation variable measurement data information in the time dimension, finally obtaining the spatiotemporal correlation graph of the substation equipment.

[0119] The beneficial effects are as follows: By retrieving electrical wiring diagrams, equipment physical layout ledgers, and cable connection lists from substations, each independent substation device is identified as a node, and the actual connection relationships between devices are represented as edges. This establishes a graph structure foundation, accurately reconstructing the spatial topology of substation equipment and providing a clear and realistic structural carrier for subsequent association of equipment operation data. By separating the timestamp information from the electromagnetic operating variable measurement dataset and associating these data with the corresponding devices in chronological order, a time-related graph is obtained. This graph structure simultaneously carries the spatial connection relationships of devices and electromagnetic operating variable measurement data at different times, effectively combining the spatial layout of equipment with the temporal dimension of operating status, facilitating the tracking of equipment operating status at different times. By extracting the weight attributes of spatial topology relationships from electrical parameter ledgers and other materials, and binding them one by one to the corresponding edges of the time-related graph, a spatiotemporal related graph is obtained. This graph structure simultaneously covers the key attribute information of equipment spatial connections and the time-dimensional operating data information, fully integrating the spatiotemporal characteristics of equipment. It can more comprehensively and accurately reflect the actual operating relationship status between devices, providing rich and reliable information support for subsequent analysis of the spatiotemporal interaction effects between devices and evaluation of the overall operating status of equipment.

[0120] S4. Based on the dielectric loss factor measurement data, perform anomaly assessment on the comprehensive electrical variable characteristic parameters to obtain the initial abnormal characteristic parameters of the substation equipment.

[0121] In this embodiment of the invention, the step of evaluating the anomalies of the comprehensive electrical variable characteristic parameters based on the dielectric loss factor measurement data to obtain the initial anomaly characteristic parameters of the substation equipment includes:

[0122] Based on the historical time-series data of the substation equipment, the reference for the change of the dielectric loss factor in the dielectric loss factor measurement data is determined;

[0123] A correlation analysis is performed on the comprehensive electrical variable characteristic parameters and the dielectric loss factor change benchmark to construct anomaly judgment rules for the comprehensive electrical variable characteristic parameters;

[0124] The deviation between the measured dielectric loss factor data and the dielectric loss factor change benchmark is compared differentially to obtain the deviation status of the measured dielectric loss factor data.

[0125] By jointly identifying the abnormal state and the deviation state in the anomaly judgment rule, a preliminary anomaly judgment conclusion for the substation equipment is obtained.

[0126] Based on the preliminary anomaly assessment, the initial anomaly level of the comprehensive electrical variable characteristic parameters is determined.

[0127] The initial anomaly judgment and the initial anomaly level are parameterized and encapsulated to obtain the initial anomaly characteristic parameters of the substation equipment.

[0128] The correlation analysis between the comprehensive electrical variable characteristic parameters and the dielectric loss factor variation benchmark, to construct anomaly judgment rules for the comprehensive electrical variable characteristic parameters, includes:

[0129] Based on the reference threshold range and trend change characteristics in the reference for dielectric loss factor variation, the reference related elements for dielectric loss of the substation equipment are determined.

[0130] According to the parameter types of voltage, current, and phase difference, the comprehensive electrical variable characteristic parameters are classified to obtain the classified electrical variable characteristic subset of the substation equipment;

[0131] Based on the correlation between the subset of classified electrical variable features and the associated elements of the dielectric loss benchmark, a parameter benchmark association mapping table for the substation equipment is constructed.

[0132] Based on the parameter benchmark association mapping table, the abnormal triggering conditions of the feature subset of the classified electrical variables are determined;

[0133] By integrating the aforementioned abnormal triggering conditions, the abnormal judgment rules for the comprehensive electrical variable characteristic parameters are obtained.

[0134] First, the historical operating sequence database of the substation equipment is accessed. This database stores the dielectric loss factor measurement data and corresponding operating condition data for all normal operating periods of the equipment over the past five years. The historical time-series data of dielectric loss factor under standard operating voltage level, normal load rate range, and normal ambient temperature and humidity conditions are filtered out from the database. Abnormal values ​​caused by equipment maintenance or external interference are removed. Then, the filtered historical data is divided into multiple time segments according to the natural year. Within each time segment, the numerical fluctuation range, mean value, and trend of dielectric loss factor over time are statistically analyzed. At the same time, the numerical deviation pattern of dielectric loss factor under different seasonal environments is analyzed. Then, the statistical results of all time segments are integrated across periods to determine the basic numerical range, allowable fluctuation range, and reasonable rate of change of dielectric loss factor under long-term normal operating conditions. These integrated numerical characteristics and trend characteristics are uniformly collected to finally obtain the dielectric loss factor change benchmark of the dielectric loss factor measurement data.

[0135] First, retrieve the established benchmark for dielectric loss factor changes in substation equipment. From this, extract two core components: the benchmark threshold range and the trend characteristics. The benchmark threshold range includes the basic value range of the dielectric loss factor under normal operating conditions, the warning value range requiring attention, and the over-limit value range prohibiting operation. The trend characteristics include the steady fluctuation trend of the dielectric loss factor during long-term stable operation, the slow upward trend when the equipment insulation is slightly deteriorated, and the abrupt upward trend when the insulation fails rapidly. Then, for each sub-range of the benchmark threshold range, analyze the corresponding equipment insulation status. For example, the basic value range corresponds to a good equipment insulation performance state, and the warning range... The numerical range corresponds to the beginning of slight insulation degradation, while the range exceeding the limit corresponds to the insulation performance being unable to guarantee safe operation. For each sub-trend of the trend change characteristics, the corresponding insulation degradation rate is analyzed. For example, a stable fluctuation trend corresponds to no obvious insulation degradation, a slow upward trend corresponds to a controllable insulation degradation rate, and a sudden upward trend corresponds to the insulation degradation entering an accelerated stage. Then, these benchmark threshold range sub-ranges, trend change characteristic sub-trends, and the meaning of the equipment operating status corresponding to each sub-range and sub-trend, which are directly related to the equipment insulation status and degradation rate, are uniformly collected to finally obtain the dielectric loss benchmark correlation elements of substation equipment.

[0136] First, retrieve the comprehensive electrical variable characteristic parameters of the generated substation equipment. Identify each characteristic item in these parameters, clarifying the core parameter type corresponding to each item. Classify all characteristic items reflecting voltage amplitude changes, voltage frequency distribution, and voltage time-domain waveform characteristics as voltage-related electrical variable characteristics; classify all characteristic items reflecting current amplitude changes, current frequency distribution, and current time-domain waveform characteristics as current-related electrical variable characteristics; and classify all characteristic items reflecting the degree of phase shift between voltage and current, phase change trends, and phase frequency response as phase difference-related electrical variable characteristics. Then, analyze the characteristic items in each category... The format is standardized by adding a voltage-specific identifier to each feature item of voltage-type electrical variable features, a current-specific identifier to each feature item of current-type electrical variable features, and a phase-difference-specific identifier to each feature item of phase-difference-type electrical variable features. At the same time, an independent feature list is established for each type of feature, which includes the identifier of the feature item, the meaning of the feature, and the data source. After completing the classification and standardization of all feature items, the voltage-type electrical variable features, current-type electrical variable features, and phase-difference-type electrical variable features are integrated into independent feature sets, thereby obtaining the substation equipment classified electrical variable feature subsets.

[0137] First, retrieve the characteristic subsets of categorized electrical variables of substation equipment and their correlation elements with dielectric loss benchmarks. For each characteristic subset of categorized electrical variables, analyze the correlation between its internal characteristic items and dielectric loss benchmark correlation elements. For example, in the voltage category, when the voltage amplitude anomaly characteristic item exceeds the normal range, the corresponding dielectric loss benchmark correlation element will move from the basic value range to the warning value range. In the current category, when the current harmonic distortion characteristic item increases in the current category, the corresponding dielectric loss benchmark correlation element will change from a stable fluctuation trend to a slow upward trend. For phase difference type electrical variables... When the phase offset abrupt change feature item in the feature subset increases sharply, the corresponding dielectric loss reference correlation element will change from a slow upward trend to an abrupt upward trend. Then, the correspondence between all feature items and dielectric loss reference correlation elements is sorted out, and a structured table is established that includes the category of the electrical variable feature subset, the specific feature item in the feature subset, the corresponding dielectric loss reference correlation element, and the correlation logic description. The table needs to clearly define what specific change in the dielectric loss reference correlation element will be triggered by the numerical change of each feature item. After completing the table content filling and logic verification, the parameter reference correlation mapping table of substation equipment is finally obtained.

[0138] First, retrieve the parameter benchmark association mapping table of the substation equipment that has been constructed. For each category of the classified electrical variable feature subset in the mapping table, extract the correspondence between the feature items under that category and the dielectric loss benchmark association elements. For the voltage-type electrical variable feature subset, based on the association logic between its feature items and dielectric loss benchmark association elements, determine that when the value of the voltage amplitude abnormal feature item exceeds the preset range and the corresponding dielectric loss benchmark association element enters the warning value range, a voltage-type electrical variable feature abnormality is triggered. When the fluctuation amplitude of the voltage frequency offset feature item exceeds the limit value and the corresponding dielectric loss benchmark association element turns into a slow upward trend, a voltage-type deep abnormality is triggered. For the current-type electrical variable feature subset, determine that when the distortion degree of the current harmonic distortion feature item reaches the threshold... When the value of the current amplitude suddenly increases and the corresponding dielectric loss benchmark related element enters the warning value range, an anomaly in the current-type electrical variable characteristic is triggered. When the value of the current amplitude suddenly increases and the corresponding dielectric loss benchmark related element turns into a sudden upward trend, a current-type deep anomaly is triggered. For the phase difference type electrical variable characteristic subset, it is clear that when the value of the phase shift degree characteristic item exceeds the basic range and the corresponding dielectric loss benchmark related element enters the warning value range, a phase difference type electrical variable characteristic anomaly is triggered. When the phase shift sudden change characteristic item appears and the corresponding dielectric loss benchmark related element turns into a sudden upward trend, a phase difference type deep anomaly is triggered. All the triggering conditions formulated for different categories of electrical variable characteristic subsets are collected to determine the abnormal triggering conditions of the category electrical variable characteristic subset.

[0139] First, retrieve the abnormal triggering conditions of the determined subset of categorized electrical variable features. Logically analyze all triggering conditions, grouping them by category: voltage, current, and phase difference triggering conditions. Then, analyze the correlation logic between different categories of triggering conditions. Clarify that when a single category triggers a basic anomaly and other categories have no anomalies, it is judged as a local parameter anomaly; when two or more categories trigger basic anomalies simultaneously, it is judged as a multi-parameter coordinated anomaly; and when any category triggers a deep anomaly, it is judged as a severe performance anomaly. Subsequently, integrate these logical correlation rules with the specific triggering conditions of each category to form a complete judgment system that includes the specific content of the triggering conditions, the combination logic of the triggering conditions, and the corresponding anomaly types. This system needs to clearly define the anomaly level and anomaly type of the comprehensive electrical variable feature parameters under different triggering conditions. After completing the logical closed-loop verification of the system, the anomaly judgment rules for the comprehensive electrical variable feature parameters are finally obtained.

[0140] First, the deviation of the real-time value of the dielectric loss factor measurement data from the dielectric loss factor change benchmark is calculated. Specifically, the real-time measurement value is compared with the basic value range in the benchmark to determine the deviation of the real-time value from the benchmark mean. Then, the deviation values ​​at fixed time intervals are extracted for multiple consecutive moments. The deviation values ​​at adjacent moments are compared to analyze the direction and magnitude of the deviation. If the deviation value continues to increase and exceeds the allowable fluctuation range of the benchmark, it is determined to be a continuously deteriorating deviation state. If the deviation value fluctuates slightly within the allowable fluctuation range, it is determined to be a slight fluctuation deviation state. If the deviation value suddenly jumps to outside the allowable range of the benchmark, it is determined to be a sudden over-limit deviation state. The deviation comparison results of all moments are integrated with the corresponding state judgment to finally obtain the deviation state of the dielectric loss factor measurement data.

[0141] First, the anomaly judgment rules for the established comprehensive electrical variable characteristic parameters and the deviation status of the obtained dielectric loss factor measurement data are retrieved. Each anomaly status specified in the anomaly judgment rules is broken down, and the current value of each comprehensive electrical variable characteristic parameter is substituted into the anomaly judgment rules for matching to determine the anomaly status type corresponding to the comprehensive electrical variable characteristic parameter. Then, this anomaly status type is cross-verified with the deviation status of the dielectric loss factor measurement data. If the comprehensive electrical variable characteristic parameter is determined to be without anomaly and the dielectric loss factor deviation status is a slight fluctuation, the substation equipment is judged to be operating normally. If the comprehensive electrical variable characteristic parameter is determined to be slightly abnormal and the dielectric loss factor deviation status is a continuously deteriorating deviation, the equipment is judged to have a potential risk of insulation degradation. If the comprehensive electrical variable characteristic parameter is determined to be severely abnormal and the dielectric loss factor deviation status is a sudden over-limit deviation, the equipment is judged to have a significant potential fault. All cross-verification results are summarized and judged to finally obtain the preliminary anomaly judgment conclusion for the substation equipment.

[0142] First, the initial anomaly assessment conclusion of the substation equipment is retrieved, and simultaneously, a historical fault case database of similar substation equipment is accessed. This database stores data on the severity and impact range of equipment faults corresponding to different anomaly conclusions. The initial anomaly assessment conclusion of the current equipment is matched with similar conclusions in the historical fault case database. The fault development stage, the potential equipment damage, and the impact level on power grid operation corresponding to the initial anomaly assessment conclusion are analyzed. Then, based on a preset level classification standard, which takes the degree of impact of the anomaly on equipment operation as the core criterion, the state corresponding to the initial anomaly assessment conclusion is divided into different levels. If the anomaly only affects local equipment parameters and there is no risk of fault development, it is classified as the lowest level. If the anomaly has caused a decline in the core performance of the equipment and there is a potential for fault expansion, it is classified as the highest level. After completing the level matching and judgment, the initial anomaly level of the comprehensive electrical variable characteristic parameters is obtained.

[0143] First, the initial anomaly assessment conclusions of substation equipment are parameterized and converted, and a unique status identifier is set for different types of anomaly assessment conclusions. This identifier can accurately correspond to the core content of the conclusion. Then, the initial anomaly level is numerically assigned, with different levels corresponding to fixed level values. Subsequently, the status identifiers and level values ​​are associated and bound together. At the same time, auxiliary information such as the judgment basis corresponding to the initial anomaly assessment conclusion and the classification source corresponding to the initial anomaly level is added. These converted parameters, assignments, and auxiliary information are integrated in a unified data format to form a structured parameter set. This parameter set contains the core identifier of the anomaly state, the level value, and the judgment traceability information, and finally, the initial anomaly characteristic parameters of the substation equipment are obtained.

[0144] The beneficial effects include establishing a benchmark for dielectric loss factor changes based on historical time-series data of substation equipment, providing a reference that aligns with the actual operating patterns of the equipment for subsequent anomaly assessment, and ensuring the rationality and relevance of anomaly detection. By conducting correlation analysis between comprehensive electrical variable characteristic parameters and the dielectric loss factor change benchmark, and constructing anomaly judgment rules, a correlation judgment logic between comprehensive electrical variable characteristic parameters and equipment insulation status was established, providing a clear and practical judgment standard for equipment anomaly identification. Differential comparison of the deviation between the dielectric loss factor measurement data and the change benchmark yields the deviation status, accurately capturing the dynamic changes in the dielectric loss factor and promptly detecting abnormal fluctuations and deterioration trends. Joint identification of the abnormal status and dielectric loss deviation status in the anomaly judgment rules yields an initial anomaly judgment conclusion, achieving cross-verification of multi-dimensional status information, significantly improving the accuracy and reliability of the initial judgment of equipment anomaly status, and effectively avoiding misjudgments and omissions due to single-dimensional judgment. Classifying the initial anomaly level of comprehensive electrical variable characteristic parameters based on the initial anomaly judgment conclusion clarifies the severity and potential impact of equipment anomalies, providing a graded decision-making basis for subsequent operation and maintenance. By parameterizing and encapsulating the initial anomaly judgment and initial anomaly level to obtain initial anomaly feature parameters, the anomaly information is structured and standardized, which facilitates the rapid identification, storage and retrieval of equipment anomaly status by the subsequent system, and provides efficient data support for intelligent operation and maintenance and fault early warning of substation equipment.

[0145] S5. Combining the functional load data of the substation equipment, and through electrothermal coupling characteristic analysis, the initial abnormal characteristic parameters are reconstructed into the fault status monitoring information of the substation equipment.

[0146] In this embodiment of the invention, the step of combining the functional load data of the substation equipment and reconstructing the initial abnormal characteristic parameters into fault status monitoring information of the substation equipment through electrothermal coupling characteristic analysis includes:

[0147] The current load type, load duration, and load fluctuation amplitude of the substation equipment are collected to obtain the load characteristic dataset of the substation equipment;

[0148] The abnormal parameter type and abnormality degree in the initial abnormal feature parameters are taken as the core abnormality elements of the substation equipment.

[0149] Electrothermal coupling characteristic analysis is performed on the load feature dataset and the abnormal core elements to establish the correlation and correspondence between the load feature dataset and the abnormal core elements;

[0150] Based on the aforementioned correlation relationship, the equipment failure cause corresponding to the initial abnormal feature parameter is determined, and the failure cause determination result of the substation equipment is obtained.

[0151] By combining the initial abnormal feature parameters, the fault cause determination results, and the load feature dataset heterogeneously, the fault status monitoring information of the substation equipment is obtained.

[0152] First, connect to the substation equipment's real-time operating condition monitoring system, load operation log, and on-site maintenance data acquisition terminal. Retrieve the load circuit type identifier currently connected to the equipment from the real-time operating condition monitoring system to determine the current load type of the equipment. The load type needs to be distinguished as inductive, capacitive, resistive, or mixed loads. Simultaneously, read the cumulative running time of the current load from startup to the data acquisition time using the system's built-in timing module to determine the load duration. Then, extract the maximum and minimum load power values ​​within the most recent acquisition period from the system's load power monitoring module. By comparing the difference between the two, determine the load fluctuation amplitude. Finally, retrieve the recent data from the load operation log for the equipment. Load switching records are cross-verified against the load types collected by the real-time operating condition monitoring system to avoid misclassification due to incorrect circuit identification. The actual load access status of the equipment is obtained through the on-site operation and maintenance acquisition terminal. The timing data of the load duration is calibrated, and the calculation results of the load fluctuation amplitude are reviewed. Abnormal fluctuation data caused by monitoring module failure are removed. After all data collection and verification are completed, the calibrated current load type load duration and load fluctuation amplitude are integrated according to a unified operating condition data format. Operating condition feature identifiers and collection timestamps are added to each data item to finally obtain the load characteristic dataset of the substation equipment.

[0153] First, retrieve the initial abnormal characteristic parameters of the generated substation equipment. From these parameters, extract two core information categories: abnormal parameter type and abnormality degree. The abnormal parameter type must be clearly identified as an abnormality of voltage, current, phase difference, or dielectric loss. Also, label the specific characteristic items corresponding to each type of abnormal parameter. The abnormality degree corresponds to the initial abnormality level in the initial abnormal characteristic parameters, which must be clearly defined as mild, moderate, or severe. Add a unique abnormality type identifier to each parameter type and a corresponding degree quantification label to each degree level. Then, associate and bind the two types of information to ensure that each abnormal parameter type corresponds to a unique abnormality degree label. Subsequently, establish an element index for the integrated information. The index includes the abnormal parameter type identifier, the corresponding characteristic item, the abnormality degree label, and the judgment source. After completing the above standardization and index establishment, uniformly collect the associated and indexed abnormal parameter types and abnormalities to form the core abnormal elements of the substation equipment.

[0154] First, an electrothermal coupling characteristic analysis framework is established. This framework focuses on the correlation logic between the electrothermal effect generated by the equipment load and the abnormal electromagnetic parameters of the equipment. It first decomposes the load duration and load fluctuation amplitude of the current load type in the load characteristic dataset into dimensions, and then decomposes the abnormal parameter types and degrees from the core abnormal elements into features. Subsequently, coupling analysis is carried out for different load types. For example, when the equipment load type is inductive, the analysis examines the electromagnetic induction heating effect of the reactive power generated during the operation of this load on the equipment windings, and the correlation logic between this effect and abnormal dielectric loss parameters. For different load durations, the analysis examines the cumulative effect of equipment temperature caused by long-term load operation and the correlation logic between voltage and other electrical variables. Anomalies are identified through correlations, such as when the load duration exceeds the normal threshold, the internal temperature rise of the equipment can cause the insulation component parameters to deviate, leading to voltage waveform distortion. For different load fluctuation amplitudes, the correlation logic between the equipment current surges caused by frequent load fluctuations and anomalies in current-related electrical variables is analyzed. For example, when the load fluctuation amplitude is large, frequent changes in current can cause poor contact of the equipment contacts, leading to abnormal current amplitude. After completing the single-dimensional coupling analysis, multi-dimensional cross-analysis is carried out to clarify the types and degrees of abnormal core elements corresponding to different load characteristic combinations. Finally, all the correlation logic obtained from the analysis is sorted out and integrated to establish a one-to-one correspondence between the load characteristic dataset and the abnormal core elements.

[0155] First, retrieve the established correlation between the load characteristic dataset and the core abnormal elements. Then, extract the specific content of the current substation equipment's load characteristic dataset and the core abnormal elements. Substitute the current load type, load duration, and load fluctuation amplitude from the load characteristic dataset into the correlation relationship to match the corresponding abnormal parameter types and degrees. Subsequently, combine the equipment's historical fault cases and electrothermal coupling characteristic analysis logic to analyze the root cause of the equipment anomaly under this combination of characteristics. For example, when the load characteristic is a long-term continuous operation of an inductive load with a large load fluctuation amplitude, and the corresponding core abnormal element is a moderate anomaly of dielectric loss, it can be determined that the fault cause is the combined effect of the continuous temperature rise caused by the long-term operation of the inductive load and the current surge caused by the load fluctuation, resulting in slight deterioration of the equipment's insulation components. When the load characteristic is a short-term high-amplitude fluctuation of a capacitive load, and the corresponding core abnormal element is a slight voltage anomaly, it can be determined that the fault cause is the voltage resonance caused by the capacitive load fluctuation, resulting in a slight distortion of the voltage waveform. After completing all matching and analysis, the determined fault causes are standardized and described to clarify the type, scope of influence, and generation mechanism of the causes, ultimately obtaining the fault cause determination result of the substation equipment.

[0156] First, the heterogeneous data formats of the initial abnormal characteristic parameter fault cause determination results and load characteristic dataset are unified. Fault diagnosis association identifiers are added to the initial abnormal characteristic parameters, operating condition association identifiers are added to the fault cause determination results, and abnormal association identifiers are added to the load characteristic dataset. Then, time-series alignment is performed based on the collection timestamps of the three types of data to ensure that the three types of data at the same monitoring time can be accurately matched. Subsequently, a unified heterogeneous data association index is established, which contains the identifier of the data source corresponding time and associated data entries for each data. Then, basic equipment information is added to the integrated data, including the equipment model, operating years, and installation location. At the same time, the data collection and verification process description is supplemented to ensure data traceability. Finally, the initial abnormal characteristic parameter fault cause determination results and load characteristic datasets with unified format, time-series alignment, and indexing are collected as a whole to form a structured information set containing equipment operating condition data, abnormal parameter data, and fault cause data. This set can comprehensively reflect the current abnormal operating status and fault root causes of the equipment, and finally obtain the fault status monitoring information of the substation equipment.

[0157] The beneficial effects include: by integrating multiple monitoring and record-keeping systems and combining them with on-site verification, the system collects the current load type, load duration, and load fluctuation amplitude of substation equipment and integrates them into a load characteristic dataset. This provides core data on equipment operating conditions, offering a precise and comprehensive foundation for subsequent fault cause analysis. By decomposing and associating abnormal parameter types and degrees from initial abnormal characteristic parameters, core abnormal elements are formed, key information dimensions of equipment anomalies are condensed, and the core direction of the anomalies is clarified, providing a clear analytical object for subsequent correlation analysis between operating conditions and anomalies. By building an electrothermal coupling characteristic analysis framework, single-dimensional and multi-dimensional cross-analysis of load characteristics and core abnormal elements is conducted, establishing a correlation between the two and connecting the inherent logical link between equipment load conditions and electromagnetic parameter anomalies, providing a scientific basis for fault cause determination. Based on the established correlation, combined with historical equipment fault cases and electrothermal coupling logic, the fault causes corresponding to initial abnormal characteristic parameters can be determined, accurately locating the root cause of equipment anomalies, effectively avoiding the blindness of traditional fault investigation, and improving the efficiency and accuracy of fault tracing. By unifying the format, aligning the time sequence, and establishing the index for the initial abnormal characteristic parameters, fault cause determination results, and load characteristic datasets, heterogeneous data are combined to form fault status monitoring information. This achieves the organic integration of multi-dimensional equipment operation data, providing comprehensive and structured information support for fault handling, operation and maintenance decisions, and status early warning of substation equipment.

[0158] S6. The spatiotemporal correlation diagram is associated and merged with the fault status monitoring information to form a status monitoring report of the substation equipment.

[0159] In this embodiment of the invention, the step of associating and fusing the spatiotemporal correlation diagram with the fault status monitoring information to form a status monitoring report for the substation equipment includes:

[0160] By analyzing the device node identifiers, node electrical parameter data, and topological association weights between devices in the spatiotemporal correlation diagram, the spatiotemporal monitoring element set of the substation equipment is obtained.

[0161] The fault status monitoring information is structured and decomposed to obtain the core fault information of the substation equipment;

[0162] By binding the spatiotemporal monitoring element set and the core fault information, the spatiotemporal fault association dataset of the substation equipment is obtained;

[0163] According to the equipment area and fault level of the substation equipment, the spatiotemporal fault association dataset is sorted spatiotemporally to obtain the spatiotemporal fault association data sequence of the substation equipment.

[0164] The spatiotemporal fault-related data sequence is deconstructed and compiled to form a status monitoring report of the substation equipment.

[0165] First, retrieve the spatiotemporal relationship diagram of the substation equipment that has already been constructed. This diagram contains equipment nodes, connection edges, and corresponding binding information. Identify each equipment node in the diagram one by one, extracting a unique equipment node identifier for each node. This identifier contains the equipment type and deployment location attributes. Simultaneously, retrieve the bound electromagnetic operating variable measurement data from each node. This data includes voltage waveform measurement data, current waveform measurement data, and dielectric loss factor measurement data corresponding to different timestamps. Classify this data into node electrical parameter data. Then, analyze each connection edge in the diagram, extracting the spatial topology relationship weight attributes bound to each edge. These attributes include the rated current carrying capacity, electrical distance, and inherent resistance of the connecting cables between equipment. The resistance level and mechanical bearing capacity are uniformly categorized as the topology association weights between equipment. Then, the extracted equipment node identifiers are format-validated to ensure that the attribute information of each identifier is complete and unambiguous. The node electrical parameter data is time-series calibrated to ensure that the timestamps of all data maintain a unified time base. The topology association weights between equipment are range-verified to remove values ​​that exceed the normal range due to input errors. After completing all verifications and calibrations, the equipment node identifiers, node electrical parameter data, and topology association weights between equipment are associated and bound according to the equipment corresponding to the node. A unique element index is established for each set of association information. The index contains the source of the information and the corresponding association relationship, and finally, the spatiotemporal monitoring element set of substation equipment is obtained.

[0166] First, retrieve the generated fault status monitoring information of the substation equipment. This information includes three types of heterogeneous data: initial abnormal characteristic parameters, fault cause determination results, and load characteristic datasets. The monitoring information is then structured and decomposed. From the initial abnormal characteristic parameters, extract the abnormal parameter types, abnormality levels, and corresponding judgment criteria. From the fault cause determination results, extract the fault cause types, impact ranges, and generation mechanisms. From the load characteristic dataset, extract the current load type, load duration, and load fluctuation amplitude. Then, classify and organize all the decomposed information. Abnormal parameter types and abnormality levels are categorized as abnormal status information; fault cause types and generation mechanisms are categorized as fault root cause information; load characteristic related data are categorized as operating condition background information; and the judgment criteria and impact ranges corresponding to each type of information are categorized as auxiliary evidence information. Then, add unique information tags to each type of information and establish an association index between the information to ensure that abnormal status information can be mapped to unique fault root cause information and operating condition background information. After completing the organization and indexing, the categorized and integrated core information is uniformly collected to obtain the core fault information of the substation equipment.

[0167] First, the spatiotemporal monitoring element set and the core fault information are adapted to a unified format. A fault association identifier is added to each device node identifier in the spatiotemporal monitoring element set, and a device matching identifier is added to each type of information in the core fault information. Then, based on the correspondence between the device matching identifier and the device node identifier, the content in the core fault information is bound one by one to the corresponding device node in the spatiotemporal monitoring element set. For example, the abnormal status information of a certain device is bound to its corresponding device node identifier, the fault root cause information of the device is bound to the association position of its node electrical parameter data, and the operating condition background information is bound to the device topology association weight corresponding to the node. At the same time, a cross-dataset association verification mechanism is established to verify whether there are logical conflicts in the bound information. For example, when the electrical parameter data of a device node has no obvious abnormality, its corresponding abnormal status information should be marked as no abnormality. If a conflict occurs, the data source is traced back and corrected. After all binding and verification are completed, the bound and integrated information is classified according to the distribution of device nodes to form a unified dataset containing equipment spatiotemporal monitoring information and core fault information, namely, the spatiotemporal fault association dataset of substation equipment.

[0168] First, the equipment deployment ledger of the substation is retrieved to clarify the physical deployment area corresponding to each piece of equipment. The entire substation is divided into equipment areas such as the high-voltage incoming line area, the main transformer area, the low-voltage outgoing line area, and the reactive power compensation area. At the same time, the fault level corresponding to the degree of abnormality in the core fault information is identified, which includes three levels: minor abnormality, moderate abnormality, and severe abnormality. Then, the spatiotemporal fault correlation dataset is grouped. According to the equipment area, all information in the dataset is divided into different regional data groups. Each group corresponds to all equipment-related information in a specific equipment area. Then, within each regional data group, the information is sorted in descending order of fault level, with severe abnormality information prioritized, followed by moderate abnormality information, and minor abnormality information last. At the same time, within the same fault level, the information is sorted a second time according to the timestamp corresponding to the information from earliest to latest to ensure the temporal continuity of the information. After completing all grouping and sorting, the sorting results of each region are arranged as a whole according to the importance of the equipment area, with the high-voltage incoming line area and the main transformer area group presented first. Finally, the spatiotemporal fault correlation data sequence of the substation equipment is obtained.

[0169] First, the generated spatiotemporal fault correlation data sequence of substation equipment is deconstructed, and the grouping logic and sorting rules of the sequence are analyzed to clarify the equipment area, fault level, and time sequence information corresponding to each data entry. Then, the deconstructed information is classified and compiled according to the standard modules of the status monitoring report. First, the equipment area overview module is compiled, which integrates the equipment node identifiers and corresponding basic operating parameters of each equipment area, clearly presenting the equipment distribution and basic operating status of each area. Next, the fault status module is compiled, which extracts abnormal status information and fault root cause information of each area, detailing the abnormal type, severity, and root cause of each equipment. Then, the spatiotemporal correlation analysis module is compiled, which combines the topological correlation weights between equipment to analyze the potential impact of faulty equipment on surrounding related equipment, presenting the spatiotemporal transmission risk of faults. Finally, the operation and maintenance handling suggestion module is compiled, which combines load characteristic data and fault causes to formulate corresponding operation and maintenance inspection and handling measures for different fault levels and types. After all modules are compiled, the report content is logically verified to ensure that the information in each module is conflict-free and complete, and that the language is accurate and conforms to the reading habits of operation and maintenance personnel, ultimately forming a status monitoring report for substation equipment.

[0170] The beneficial effects are as follows: By analyzing the device node identifiers, node electrical parameter data, and topological association weights between devices in the spatiotemporal correlation diagram and completing verification and calibration, a spatiotemporal monitoring element set is obtained. This set integrates the spatial topology information, multi-time-dimensional electrical parameter data, and topological association attributes of substation equipment, providing a comprehensive and accurate spatiotemporal monitoring foundation for subsequent fault correlation analysis. Through structured decomposition and classification of fault status monitoring information, core fault information is obtained, condensing key dimensions such as abnormal equipment states, fault root causes, and operating conditions. Redundant data is eliminated, and the correlation logic between various pieces of information is clarified, providing a clear core object for binding faults with spatiotemporal monitoring data. By adapting the format and binding the spatiotemporal monitoring element set with the core fault information, a spatiotemporal fault correlation dataset is obtained. This achieves the organic integration of equipment spatiotemporal monitoring data and core fault information, establishing a link between equipment operating status and fault information, facilitating the tracing of the spatiotemporal operating characteristics of the equipment corresponding to the fault. The spatiotemporal fault correlation dataset is spatiotemporally sorted according to equipment region and fault level to obtain a spatiotemporal fault correlation data sequence. This gives the data a clear regional classification and fault priority, facilitating the rapid location of high-risk areas and serious faults while ensuring the temporal continuity of the data, thus improving readability and usability. By deconstructing and compiling the spatiotemporal fault correlation data sequence to generate a status monitoring report, the scattered correlation data is transformed into a systematic report containing equipment region overview, fault status, spatiotemporal propagation risks, and operation and maintenance handling suggestions. This provides intuitive and comprehensive information support for substation equipment operation and maintenance decisions, fault handling, and status early warning, significantly improving the efficiency and accuracy of equipment status management.

[0171] like Figure 2 The diagram shown is a functional block diagram of a substation equipment operation data monitoring system provided in an embodiment of the present invention.

[0172] The substation equipment operation data monitoring system 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the substation equipment operation data monitoring system 100 may include a data acquisition module 101, a time-frequency domain analysis module 102, a spatial correlation graph construction module 103, an anomaly assessment module 104, an electrothermal coupling analysis module 105, and a monitoring report generation module 106. The modules described in this invention can also be referred to as units, which are a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, stored in the memory of the electronic device.

[0173] In this embodiment, the functions of each module / unit are as follows:

[0174] The data acquisition module 101 is used to integrate the voltage waveform measurement data, current waveform measurement data and dielectric loss factor measurement data of the substation equipment into the electromagnetic operating variable measurement dataset of the substation equipment.

[0175] The time-frequency domain analysis module 102 is used to perform time-frequency feature analysis on the electromagnetic operating variable measurement dataset to obtain the comprehensive electrical variable feature parameters of the substation equipment.

[0176] The spatial association graph construction module 103 is used to construct a spatiotemporal association graph of the substation equipment based on the timestamp information of the electromagnetic operating variable measurement dataset and the spatial topology relationship of the substation equipment.

[0177] The anomaly assessment module 104 is used to assess the anomalies of the comprehensive electrical variable characteristic parameters based on the dielectric loss factor measurement data, and to obtain the initial anomaly characteristic parameters of the substation equipment.

[0178] The electrothermal coupling analysis module 105 is used to combine the functional load data of the substation equipment and reconstruct the initial abnormal characteristic parameters into the fault status monitoring information of the substation equipment through electrothermal coupling characteristic analysis.

[0179] The monitoring report generation module 106 is used to associate and merge the spatiotemporal correlation diagram with the fault status monitoring information into a status monitoring report of the substation equipment.

[0180] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0181] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0182] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0183] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0184] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0185] Finally, 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.

Claims

1. A method of monitoring substation equipment operation data, characterized by, The method comprises: S1, integrating voltage waveform measurement data, current waveform measurement data and dielectric loss factor measurement data of a substation equipment into an electromagnetic operating variable measurement data set of the substation equipment; S2, performing time-frequency feature analysis on the electromagnetic operating variable measurement data set to obtain comprehensive electrical variable feature parameters of the substation equipment, comprising: performing time-frequency domain joint analysis on the voltage waveform measurement data and the current waveform measurement data in the electromagnetic operating variable measurement data set to obtain a voltage time-frequency feature matrix and a current time-frequency feature matrix of the substation equipment; based on the voltage time-frequency feature matrix and the current time-frequency feature matrix, evaluating the phase difference between voltage and current in the substation equipment to obtain a phase difference candidate feature parameter of the substation equipment; taking the frequency domain variation feature of the dielectric loss factor measurement data in the electromagnetic operating variable measurement data set as a dielectric loss candidate feature parameter of the substation equipment; eliminating redundant feature parameters in the phase difference candidate feature parameter and the dielectric loss candidate feature parameter to obtain core candidate feature parameters of the substation equipment, comprising: integrating the phase difference candidate feature parameter and the dielectric loss candidate feature parameter into a candidate feature parameter set of the substation equipment; obtaining the numerical distribution range of the candidate feature parameter set under the historical normal operating state and the typical fault mode of the substation equipment and the current equipment operating state; according to the numerical distribution range and the current equipment operating state, calculating the feature saliency strength of the candidate feature parameter set, wherein the calculation formula of the feature saliency strength is as follows: ； In the formula, represents the feature saliency strength of the candidate feature parameter in the candidate feature parameter set, represents the current monitoring value of the candidate feature parameter in the candidate feature parameter set, represents the mean value of the candidate feature parameter in the candidate feature parameter set under the historical normal operation state, represents the standard deviation of the candidate feature parameter in the candidate feature parameter set under the historical normal operation state, represents a preset weight coefficient for representing that the current monitoring value deviates from the normal historical interval, represents the mean value of the candidate feature parameter in the candidate feature parameter set under the typical fault mode, represents the standard deviation of the candidate feature parameter in the candidate feature parameter set under the typical fault mode, represents a function for finding the maximum value in a group of numerical values, represents a preset weight coefficient for representing that the current monitoring value approaches the fault interval.​​​​​​ judging the candidate feature parameter in the candidate feature parameter set whose feature saliency strength is lower than a preset saliency threshold as a redundant feature parameter; eliminating the redundant feature parameter from the candidate feature parameter set to obtain the core candidate feature parameter of the substation equipment; performing dimension reduction fusion on the core candidate feature parameter to obtain the comprehensive electrical variable feature parameter of the substation equipment; S3, constructing a space-time correlation graph of the substation equipment according to the timestamp information of the electromagnetic operating variable measurement data set and the spatial topology relationship of the substation equipment; S4, based on the dielectric loss factor measurement data, performing abnormality evaluation on the comprehensive electrical variable feature parameter to obtain initial abnormal feature parameters of the substation equipment, comprising: determining a dielectric loss factor change reference of the dielectric loss factor measurement data according to the historical time series data of the substation equipment; performing correlation analysis on the comprehensive electrical variable feature parameter and the dielectric loss factor change reference to construct an abnormality judgment rule of the comprehensive electrical variable feature parameter; differentially comparing the deviation degree between the dielectric loss factor measurement data and the dielectric loss factor change reference to obtain the deviation state of the dielectric loss factor measurement data; jointly identifying the abnormal state in the abnormality judgment rule and the deviation state to obtain an abnormal preliminary judgment conclusion of the substation equipment; dividing the initial abnormality level of the comprehensive electrical variable feature parameter according to the abnormal preliminary judgment conclusion; Parameterize and encapsulate the initial abnormality judgment conclusion and the initial abnormality level to obtain an initial abnormality characteristic parameter of the substation equipment; S5, in combination with the functional load data of the substation equipment, the initial abnormality characteristic parameter is reconstructed into the fault state monitoring information of the substation equipment through electric-thermal coupling characteristic analysis, including: Collecting the current load type, load duration and load fluctuation amplitude of the substation equipment to obtain a load characteristic data set of the substation equipment; Taking the abnormal parameter type and abnormal degree in the initial abnormality characteristic parameter as the abnormal core elements of the substation equipment; Performing electric-thermal coupling characteristic analysis on the load characteristic data set and the abnormal core elements to establish the corresponding relationship between the load characteristic data set and the abnormal core elements; Based on the corresponding relationship, the device fault inducement corresponding to the initial abnormality characteristic parameter is determined to obtain a fault inducement determination result of the substation equipment; Performing heterogeneous data joint on the initial abnormality characteristic parameter, the fault inducement determination result and the load characteristic data set to obtain the fault state monitoring information of the substation equipment; S6, the spatio-temporal correlation graph and the fault state monitoring information are associated and fused into a state monitoring report of the substation equipment, including: Analyzing the device node identifier, node electric parameter data and device interconnection topology correlation weight in the spatio-temporal correlation graph to obtain a spatio-temporal monitoring element set of the substation equipment; Structurally splitting the fault state monitoring information to obtain a fault core information of the substation equipment; Binding the spatio-temporal monitoring element set and the fault core information to obtain a spatio-temporal fault correlation data set of the substation equipment; According to the device area and fault level of the substation equipment, the spatio-temporal fault correlation data set is spatio-temporally sorted to obtain a spatio-temporal fault correlation data sequence of the substation equipment; The spatio-temporal fault correlation data sequence is deconstructed and compiled to form the state monitoring report of the substation equipment.

2. The substation equipment operation data monitoring method of claim 1, wherein, The voltage waveform measurement data, current waveform measurement data and dielectric loss factor measurement data of the substation equipment are integrated into an electromagnetic operation variable measurement data set of the substation equipment, including: Synchronously measuring the voltage waveform measurement data, current waveform measurement data and dielectric loss factor measurement data of the substation equipment; Adding time stamps to the voltage waveform measurement data, current waveform measurement data and dielectric loss factor measurement data to obtain a time-stamped electromagnetic measurement data set of the substation equipment; Performing logical verification on the time-stamped electromagnetic measurement data set to obtain an electromagnetic operation variable measurement data set of the substation equipment.

3. The substation equipment operation data monitoring method of claim 1, wherein, According to the time stamp information of the electromagnetic operation variable measurement data set and the spatial topology relationship of the substation equipment, a spatio-temporal correlation graph of the substation equipment is constructed, including: According to the spatial topology relationship of the substation equipment, the substation equipment is taken as a node and the connection relationship between the devices in the substation equipment is taken as an edge to construct a graph structure basis of the substation equipment; mapping the electromagnetic operation variable measurement data to nodes in the graph structure foundation based on timestamp information in the electromagnetic operation variable measurement data, to obtain a time correlation graph of the power substation device; inputting weight attributes in the spatial topological relationship to edges in the time correlation graph, to obtain a space-time correlation graph of the power substation device.

4. The substation equipment operation data monitoring method of claim 1, wherein, The correlation analysis of the comprehensive electric variable characteristic parameter and the dielectric loss factor change reference includes: According to the reference threshold interval and the trend change characteristic in the dielectric loss factor change reference, the dielectric loss reference correlation element of the power substation device is determined; According to the parameter types of voltage, current and phase difference, the comprehensive electric variable characteristic parameter is classified to obtain a classified electric variable characteristic subset of the power substation device; According to the correlation relationship between the classified electric variable characteristic subset and the dielectric loss reference correlation element, a parameter reference correlation mapping table of the power substation device is constructed; Based on the parameter reference correlation mapping table, the abnormal trigger condition of the classified electric variable characteristic subset is determined; The abnormal trigger condition is integrated to obtain the abnormal judgment rule of the comprehensive electric variable characteristic parameter.

5. A substation equipment operation data monitoring system characterized by, A power substation device operation data monitoring method for realizing the power substation device operation data monitoring method of claim 1, the system comprising: A data acquisition module for integrating voltage waveform measurement data, current waveform measurement data and dielectric loss factor measurement data of a power substation device into electromagnetic operation variable measurement data sets of the power substation device; A time-frequency domain analysis module for performing time-frequency feature analysis on the electromagnetic operation variable measurement data sets to obtain comprehensive electric variable characteristic parameters of the power substation device; A space correlation graph construction module for constructing a space-time correlation graph of the power substation device according to timestamp information of the electromagnetic operation variable measurement data sets and spatial topological relationships of the power substation device; An abnormality evaluation module for performing abnormality evaluation on the comprehensive electric variable characteristic parameters based on the dielectric loss factor measurement data to obtain initial abnormal characteristic parameters of the power substation device; An electro-thermal coupling analysis module for combining functional load data of the power substation device and reconstructing the initial abnormal characteristic parameters into fault state monitoring information of the power substation device through electro-thermal coupling characteristic analysis; A monitoring report generation module for associating and fusing the space-time correlation graph and the fault state monitoring information into a state monitoring report of the power substation device.

Images