Abnormity processing method, device and equipment for transformer substation

Abstract

The embodiment of the invention provides an exception handling method, device and equipment for a transformer substation. The method comprises the following steps: acquiring time sequence data corresponding to a plurality of substation parameters of a substation; preprocessing the time sequence data corresponding to each substation parameter to obtain processed time sequence data; performing feature extraction on each piece of processed time series data to obtain features of each piece of processed time series data; for each transformer substation parameter, based on the characteristics of the processed time series data corresponding to the transformer substation parameter, performing potential anomaly identification by adopting a pre-trained anomaly identification model, and outputting the anomaly degree of the time series data; and determining the time sequence data of which the abnormal degree is greater than a preset threshold value as potential abnormal data, and giving an alarm for the data. The method is used for accurately identifying potential abnormal data in the transformer substation and realizing early warning of faults of the transformer substation, so that the operation and maintenance efficiency and the operation stability of the transformer substation are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of substation operation and maintenance, and in particular to a substation abnormality processing method, device and equipment. BACKGROUND

[0002] With the rapid advancement of smart grid construction, the operation state monitoring and fault accurate positioning capability of the substation as the core hub of the power system are directly related to the safety and reliability of the power grid.

[0003] At present, the substation has widely deployed mutual inductors, sensors and other devices to realize real-time sensing of electrical quantities (such as voltage and current) and non-electrical quantities (such as temperature and gas pressure); abnormality positioning is realized by analyzing the sensing data, thereby completing the monitoring of the operation state of the substation and the positioning of the fault.

[0004] The traditional abnormal data positioning scheme relies on manual experience to set fixed thresholds for out-of-limit detection. However, this simple single threshold-based method cannot realize early warning of potential risks, resulting in early risks being ignored, and it is difficult to meet the efficient operation and maintenance requirements of the substation. SUMMARY

[0005] The embodiments of the present application provide a substation abnormality processing method, device and equipment to accurately identify potential abnormal data in the substation, thereby improving the efficiency of substation operation and maintenance and the stability effect of operation.

[0006] In a first aspect, the embodiments of the present application provide a substation abnormality processing method, comprising:

[0007] Obtaining time series data corresponding to a plurality of substation parameters of a substation, different substation parameters being different electrical quantity parameters or non-electrical quantity parameters of the substation;

[0008] After pre-processing the time series data corresponding to each substation parameter, obtaining the pre-processed time series data;

[0009] Performing feature extraction on each pre-processed time series data to obtain the features of each pre-processed time series data;

[0010] For each substation parameter, based on the features of the pre-processed time series data corresponding to the substation parameter, a pre-trained abnormality recognition model is used for potential abnormality recognition, and the abnormality degree of the time series data is output, the abnormality recognition model being obtained by pre-training a machine learning model based on historical substation data, and the machine learning model including an isolation forest model;

[0011] The time series data with an abnormality degree greater than a preset threshold is determined as potential abnormal data, and the potential abnormal data is alarmed.

[0012] In one possible implementation, the characteristics of the processed time series data include at least one of the following: mean, variance, maximum value, minimum value, skewness, kurtosis, dominant frequency amplitude, harmonic distortion rate, outlier percentage, and volatility of the processed time series data.

[0013] In one possible implementation, the method further includes:

[0014] For each substation parameter, the processed time-series data corresponding to the substation parameter is subjected to quality optimization processing to obtain the optimized time-series data corresponding to the substation parameter. The quality optimization processing includes at least one of cleaning, data completion processing, normalization processing, dimensionless processing, and noise removal processing based on data fluctuation.

[0015] Data in the optimized time-series data corresponding to the substation parameters that exceed the parameter thresholds corresponding to the substation parameters are identified as out-of-limit abnormal data, and an alarm is issued for the out-of-limit abnormal data.

[0016] In one possible implementation, the method further includes:

[0017] Obtain the first abnormal correlation data corresponding to the over-limit abnormal data, and the second abnormal correlation data corresponding to the potential abnormal data; the first abnormal correlation data includes the first protection action signal, fault recording segment, and switch status information during the time period of the over-limit abnormality; the second abnormal correlation data includes the second protection action signal, fault recording segment, and switch status information during the time period of the potential abnormality.

[0018] Based on the first abnormal correlation data and the second abnormal correlation data, the fault type of the substation is determined;

[0019] Based on the fault type, a pre-constructed power grid topology diagram of the substation is constructed, and the associated nodes in the power grid topology diagram are respectively associated with the out-of-limit abnormal data and the potential abnormal data. The abnormal interval or abnormal equipment of the substation is determined in the power grid topology diagram.

[0020] The fault point of the substation is determined based on the abnormal interval or abnormal equipment.

[0021] In one possible implementation, the pre-constructed power grid topology map of the substation based on the fault type, wherein the associated nodes in the power grid topology map are respectively associated with the out-of-limit abnormal data and the potential abnormal data, and the abnormal intervals and abnormal equipment of the substation are determined in the power grid topology map, including:

[0022] In the pre-constructed power grid topology diagram of the substation, starting from the associated nodes associated with the out-of-limit abnormal data and the associated nodes associated with the potential abnormal data, within the fault association area matching the fault type, based on the electrical connections in the power grid topology diagram and the first protection action information and the second protection action information, a breadth-first search algorithm is used to determine the breakpoint of the substation, where the breakpoint is the node that causes the interruption of power transmission in the power grid.

[0023] Based on the electrical connection of the substation's breakpoint in the power grid topology diagram, the interval associated with the breakpoint is obtained as an abnormal interval, or the device associated with the breakpoint is obtained as an abnormal device.

[0024] In one possible implementation, determining the fault point of the substation based on the abnormal interval or abnormal equipment includes:

[0025] The abnormal intervals or abnormal equipment are characterized by traveling wave ranging and / or synchronous phasor measurement technology, and the fault point of the substation is determined based on the detection results.

[0026] In one possible implementation, the method further includes:

[0027] The substation is modeled using 3D rendering technology to obtain a model of the substation.

[0028] The time-series data of the multiple substation parameters and / or the fault points are visualized in the model.

[0029] In one possible implementation, the preprocessing includes analog-to-digital conversion, filtering, scaling, and recording event sequence processing.

[0030] Secondly, embodiments of this application provide an anomaly handling device for a substation, comprising:

[0031] The acquisition module is used to acquire time-series data corresponding to multiple substation parameters of the substation, where different substation parameters are different electrical quantity parameters or non-electrical quantity parameters of the substation.

[0032] The preprocessing module is used to preprocess the time series data corresponding to each substation parameter to obtain the processed time series data.

[0033] The feature extraction module is used to extract features from each processed time series data to obtain the features of each processed time series data.

[0034] The potential anomaly identification module identifies potential anomalies for each substation parameter based on the characteristics of the processed time-series data corresponding to the substation parameter using a pre-trained anomaly identification model and outputs the degree of anomaly of the time-series data. The anomaly identification model is pre-trained on a machine learning model based on historical substation data, and the machine learning model includes an isolated forest model.

[0035] The first determination module is used to determine time-series data with an abnormality level greater than a preset threshold as potential abnormal data and to issue an alarm for the potential abnormal data.

[0036] In one possible implementation, the features of the processed time-series data in the feature extraction module include at least one of the following: mean, variance, maximum value, minimum value, skewness, kurtosis, dominant frequency amplitude, harmonic distortion rate, outlier percentage, and volatility of the processed time-series data.

[0037] In one possible implementation, the device further includes:

[0038] The out-of-limit anomaly identification module is used to perform quality optimization processing on the processed time series data corresponding to each substation parameter to obtain the optimized time series data corresponding to the substation parameter. The quality optimization processing includes at least one of cleaning, data completion processing, normalization processing, dimensionless processing, and noise removal processing based on data fluctuation.

[0039] Data in the optimized time-series data corresponding to the substation parameters that exceed the parameter thresholds corresponding to the substation parameters are identified as out-of-limit abnormal data, and an alarm is issued for the out-of-limit abnormal data.

[0040] In one possible implementation, the apparatus further includes a second determining module, the second determining module comprising:

[0041] The acquisition unit is used to acquire first abnormal correlation data corresponding to the over-limit abnormal data and second abnormal correlation data corresponding to the potential abnormal data; the first abnormal correlation data includes a first protection action signal, fault recording segment, and switch status information during the over-limit abnormality period; the second abnormal correlation data includes a second protection action signal, fault recording segment, and switch status information during the potential abnormality period.

[0042] The first determining unit is used to determine the fault type of the substation based on the first abnormal correlation data and the second abnormal correlation data.

[0043] The second determining unit is used to pre-construct a power grid topology map of the substation based on the fault type, wherein the associated nodes in the power grid topology map are respectively associated with the out-of-limit abnormal data and the potential abnormal data, and to determine the abnormal interval or abnormal equipment of the substation in the power grid topology map.

[0044] The third determining unit is used to determine the fault point of the substation based on the abnormal interval or abnormal equipment.

[0045] In one possible implementation, the second determining unit is specifically used for:

[0046] In the pre-constructed power grid topology diagram of the substation, starting from the associated nodes associated with the out-of-limit abnormal data and the associated nodes associated with the potential abnormal data, within the fault association area matching the fault type, based on the electrical connections in the power grid topology diagram and the first protection action information and the second protection action information, a breadth-first search algorithm is used to determine the breakpoint of the substation, where the breakpoint is the node that causes the interruption of power transmission in the power grid.

[0047] Based on the electrical connection of the substation's breakpoint in the power grid topology diagram, the interval associated with the breakpoint is obtained as an abnormal interval, or the device associated with the breakpoint is obtained as an abnormal device.

[0048] In one possible implementation, the third determining unit is specifically used for:

[0049] The abnormal intervals or abnormal equipment are characterized by traveling wave ranging and / or synchronous phasor measurement technology, and the fault point of the substation is determined based on the detection results.

[0050] In one possible implementation, the device further includes a display module for:

[0051] The substation is modeled using 3D rendering technology to obtain a model of the substation.

[0052] The time-series data of the multiple substation parameters and / or the fault points are visualized in the model.

[0053] In one possible implementation, the preprocessing in the preprocessing module includes analog-to-digital conversion, filtering, scaling conversion, and recording event sequence processing.

[0054] Thirdly, embodiments of this application provide a computer device, including: a memory and a processor;

[0055] The memory stores computer-executed instructions;

[0056] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0057] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0058] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0059] The substation anomaly handling method, apparatus, and equipment provided in this application embodiment acquire time-series data corresponding to multiple substation parameters; preprocess the time-series data corresponding to each substation parameter to obtain processed time-series data; extract features from each processed time-series data to obtain features of each processed time-series data; for each substation parameter, based on the features of the processed time-series data corresponding to the substation parameter, use a pre-trained anomaly identification model to identify potential anomalies and output the anomaly degree of the time-series data; identify time-series data with an anomaly degree greater than a preset threshold as potential anomaly data and issue an alarm for that data. This method accurately identifies potential anomaly data in the substation, achieves early warning of substation faults, enables preventive measures to be taken at the fault initiation stage, effectively avoids the expansion and deterioration of faults, improves the efficiency of substation operation and maintenance, reduces power outage time and economic losses, and effectively enhances the safety, stability, and reliability of substation operation. Attached Figure Description

[0060] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0061] Figure 1 This is a flowchart illustrating the substation anomaly handling method provided in Embodiment 1 of this application;

[0062] Figure 2 This is a flowchart illustrating the substation anomaly handling method provided in Embodiment 2 of this application;

[0063] Figure 3 This is a schematic diagram of the abnormal handling device for a substation provided in Embodiment 3 of this application;

[0064] Figure 4 This is a schematic diagram of the structure of the substation fault handling device provided in Embodiment 4 of this application;

[0065] Figure 5 A schematic diagram of the structure of the computer device provided in this application.

[0066] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0067] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0068] A substation is a core node in a power system that connects power grids of different voltage levels and realizes power conversion, distribution, control, and protection. Its core function is to convert high-voltage power output from the power plant (for long-distance transmission) into medium- and low-voltage power (to meet the needs of users or downstream power grids) through transformers. It also undertakes the collection, distribution, monitoring, and fault handling of power, serving as a "hub" to ensure efficient and safe power transmission from the production end to the user end. A substation consists of multiple bays that perform specific functions such as power conversion, transmission, and distribution. Each bay contains main equipment (such as transformers, circuit breakers, and busbars) and auxiliary equipment (such as instrument transformers and protection devices) connected by lines to achieve its functions. Among these, the transformer is the key monitoring object in a substation, including critical components such as windings, core, bushings, and tap changers.

[0069] When researching methods for handling substation anomalies, the inventors considered extracting features (such as average value and distortion rate within a preset time period) of time-series data corresponding to different substation parameters (such as the current of bus A). They then used an anomaly recognition model pre-trained on a machine learning model (such as an isolated forest model). Based on the features of each parameter in the substation under normal operating conditions, which were pre-learned from historical substation data, and the features of each substation parameter extracted at present, the inventors quantified whether each substation parameter deviated from the normal range. This enabled the identification of potential anomalies based on features, thus achieving early warning of substation anomalies and minimizing the impact of faults.

[0070] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0071] Figure 1 This is a flowchart illustrating the substation anomaly handling method provided in Embodiment 1 of this application, as shown below. Figure 1 As shown, the method provided in Embodiment 1 of this application includes:

[0072] S101. Obtain the time sequence data corresponding to multiple substation parameters of the substation. The different substation parameters are different electrical quantity parameters or non-electrical quantity parameters of the substation.

[0073] In this step, substation parameters refer to parameters that reflect the operating status of different equipment or lines in the substation. These include electrical and non-electrical parameters. Electrical parameters are physical quantities directly related to the electrical characteristics of each piece of equipment or line in the substation, such as current, voltage, power factor, and harmonic content. Non-electrical parameters are physical quantities that are not directly related to the electrical characteristics of each piece of equipment or line in the substation, but can reflect the physical state and environmental conditions of each piece of equipment, such as transformer winding temperature, transformer oil level, and vibration amplitude of flexible busbars. Time-series data refers to the continuous data of substation parameters over time, used to capture the characteristics corresponding to the substation parameters.

[0074] In practical applications, to ensure real-time monitoring of substations, it is necessary to acquire time-series data corresponding to multiple substation parameters in real time, and then perform subsequent real-time preprocessing, feature extraction, and analysis on the time-series data within a preset time window.

[0075] It should be understood that different equipment or line faults will manifest themselves through their electrical and non-electrical parameters. For example, transformer windings, as core components, are prone to faults such as inter-turn short circuits, phase-to-phase short circuits, and winding grounding. During an inter-turn short circuit, the short-circuit current increases dramatically, causing magnetic flux distortion and increased harmonics, while localized overheating causes the winding temperature to spike. A phase-to-phase short circuit results in a severe short-circuit current, threatening transformer safety. Winding grounding saturates the core, increasing losses and temperature, and may also trigger protection devices to trip. These faults can be observed through parameters such as current, voltage, power factor, harmonic content, and winding temperature.

[0076] Core faults mainly include multi-point grounding of the core and short circuits between laminations. Multi-point grounding of the core creates circulating current, leading to increased grounding current and core temperature. If not addressed promptly, the core may burn out. Short circuits between laminations increase magnetic reluctance, no-load loss, and no-load current, causing localized overheating at the short-circuit location and affecting transformer performance. These faults can be detected by monitoring parameters such as grounding current, no-load loss, no-load current, and core temperature.

[0077] Bushings are responsible for connecting the transformer's internal leads to the outside of the tank, and are prone to insulation damage and flashover faults. When bushing insulation is damaged, the dielectric loss factor increases, capacitance changes, leakage current increases, and local overheating may also occur. Bushing flashovers often occur under severe weather conditions, preceded by partial discharge, resulting in increased surface temperature, which ultimately leads to transformer tripping and power outages, affecting the stable operation of the power grid. These faults can be detected by monitoring parameters such as dielectric loss factor, capacitance, leakage current, bushing surface temperature, and partial discharge signals.

[0078] Tap changers are used to adjust transformer voltage. Common faults include poor contact, burnt contacts, and misaligned tap changers. Poor contact increases contact resistance, causing voltage and current fluctuations and increased contact temperature. Burnt contacts prevent the tap changer from switching properly. Misaligned tap changers cause incorrect transformer turns ratios, affecting voltage quality and even damaging equipment. These faults can be detected by monitoring parameters such as contact resistance, voltage fluctuations, current fluctuations, and the sound and vibration of the operating mechanism.

[0079] In view of this, this solution first obtains the time-series data corresponding to different substation parameters, and then locates the substation fault by locating the abnormal data.

[0080] In one specific implementation, time-series data corresponding to multiple substation parameters can be acquired through instrument transformers and sensors. Specifically, voltage and current data of different equipment lines in the substation can be collected through instrument transformers, and temperature, humidity, and pressure data of different equipment or lines can be collected through sensors.

[0081] S102. After preprocessing the time series data corresponding to each substation parameter, the processed time series data is obtained.

[0082] In this step, to facilitate the unified processing of the data later, the directly acquired data first needs to be preprocessed.

[0083] In one possible implementation, preprocessing includes analog-to-digital conversion, filtering, scaling, and recording event sequence processing.

[0084] Specifically, analog-to-digital conversion refers to the process of converting analog signals collected by sensors into digital signals that can be recognized and processed by computers, providing a foundation for subsequent data storage, calculation, and analysis.

[0085] Filtering refers to removing noise or interference from the original data using specific algorithms (such as Gaussian filtering) or circuits, retaining valid data that reflects the true operating status of the equipment or line, and improving the accuracy of subsequent analysis results.

[0086] Scale conversion refers to converting parameter data of different physical quantities and magnitudes into engineering units with actual physical meaning (such as converting the secondary current value of a current transformer into the actual primary current value) to facilitate intuitive understanding of the data meaning.

[0087] Recording time sequence processing refers to marking the time sequence data corresponding to each collected parameter with a timestamp.

[0088] This implementation method preprocesses the time-series data corresponding to each substation parameter by performing digital-to-analog conversion, filtering, scaling conversion, and recording event sequence processing, which provides a foundation for subsequent unified data processing and analysis and ensures the accuracy of subsequent processing.

[0089] In one specific implementation, the data collected by the aforementioned transformers and sensors can be transmitted to the data acquisition terminal via a shielded cable. The data acquisition terminal preprocesses the data and uses the GOOSE / SV protocol to reduce latency. The data is then uploaded to the cloud platform for storage via an encrypted channel for subsequent steps such as feature extraction and anomaly identification.

[0090] This implementation method uploads the processed time-series data to a cloud platform for storage, aiming to centrally integrate multi-source information and build a comprehensive data foundation. The results are significant, facilitating long-term storage and easy retrieval, providing rich material for subsequent analysis, and leveraging the powerful computing capabilities and stable environment of the cloud platform to ensure data security and integrity, supporting stable system operation and decision-making.

[0091] S103. Extract features from each processed time series data to obtain the features of each processed time series data.

[0092] In this step, features are captured and extracted from each processed time-series data point to obtain the features of each processed time-series data point. These features include key information that reflects the inherent patterns and operational status of the data, such as statistical features, time-domain features, frequency-domain features, and abrupt change features.

[0093] In practical applications, it is necessary to extract the features of time series data within the current time window.

[0094] In one possible implementation, the characteristics of the processed time series data include at least one of the following: mean, variance, maximum value, minimum value, skewness, kurtosis, dominant frequency amplitude, harmonic distortion rate, outlier percentage, and volatility of the processed time series data.

[0095] Among them, the mean represents the average level of the time series data within the current time window; the variance represents the dispersion of the time series data within the current time window; the maximum and minimum values ​​represent the largest and smallest values ​​of the time series data within the current time window, respectively; the skewness represents the asymmetry of the distribution of the time series data within the current time window; the skewness also represents the concentration of the distribution of the time series data within the current time window; the dominant frequency amplitude refers to the signal strength (i.e., amplitude) corresponding to the highest energy frequency (i.e., the dominant frequency) after performing a Fourier transform on the time series data within the current time window; the harmonic distortion rate represents the degree of interference of the harmonic components of the time series data within the current time window on the fundamental frequency; the outlier ratio represents the proportion of data outside the labeled range in the time series data within the current time window; and the volatility represents the degree of fluctuation of the time series data within the current time window, which can be obtained by calculating the mean of the absolute values ​​of the differences between adjacent data.

[0096] This implementation extracts at least one feature from the processed time-series data, including mean, variance, maximum, minimum, skewness, kurtosis, dominant frequency amplitude, harmonic distortion rate, outlier percentage, and volatility. It captures the overall distribution and fluctuation characteristics of the data from a statistical perspective, intuitively reflecting the baseline state of equipment or line operation; reveals potential pattern shifts in data distribution, allowing early detection of atypical operating modes of equipment or lines; captures hidden waveform anomalies from frequency domain features, accurately locating harmonic interference caused by winding short circuits, insulation damage, etc.; and captures abrupt changes and gradual deterioration of equipment status from anomaly and trend dimensions, such as slow temperature rise due to localized overheating or increased current fluctuations due to poor contact. This transforms massive amounts of data into quantifiable and comparable key indicators, reducing the complexity of subsequent analysis and providing comprehensive feature input for the subsequent anomaly identification model, ensuring the reliability and accuracy of anomaly identification.

[0097] S104. For each substation parameter, based on the characteristics of the processed time series data corresponding to the substation parameter, a pre-trained anomaly identification model is used to identify potential anomalies and output the degree of anomaly in the time series data.

[0098] The anomaly identification model is obtained by pre-training a machine learning model based on historical substation data. The machine learning model includes the isolated forest model.

[0099] In this step, based on the characteristics corresponding to each substation parameter, a pre-trained anomaly identification model is used to identify potential anomalies. This allows the anomaly identification model to perform feature analysis on the current degree of anomaly in the time series data based on the pre-learned normal characteristics of the substation parameters and the characteristics of the current input.

[0100] It should be understood that Isolation Forest is an unsupervised anomaly detection algorithm based on a tree structure. Its core logic is to utilize the characteristic that "the feature patterns of anomalous data are significantly different from those of normal data, and therefore it is easier to quickly 'isolate' them." In actual training, the model recursively divides the data space based on the features corresponding to normal data in historical substation data, and constructs multiple isolation trees by randomly selecting features and partitioning thresholds, thereby obtaining a pre-trained anomaly recognition model.

[0101] When using anomaly detection models for anomaly detection, the features of normal data exhibit a stable distribution pattern in historical operation, and their feature combinations are more dense in the data space, requiring more partitioning (i.e., deeper path depth) to isolate them; while the feature combinations of potentially anomalous data deviate from the normal pattern and are more sparse in the data space, often requiring fewer partitioning (i.e., shallower path depth) to separate them individually. Therefore, the anomaly detection model will output the degree of anomaly of the time series data according to the depth of the tree in which the features are partitioned.

[0102] Optionally, the machine learning model may also include a Long Short-Term Memory (LSTM) network model. Specifically, historical substation data can be used to train the LSTM model, enabling it to capture temporal features in the time-series data, resulting in a trained LSTM model. In practical applications, the time-series data to be detected is input into the trained LSTM model, which then "reconstructs" the data based on the learned normal patterns. If the data to be detected is normal, its variation pattern is consistent with the normal pattern learned by the model, and the reconstruction error is small. However, if the data is abnormal (such as a sudden increase in current or temperature caused by a fault), its variation pattern deviates from the normal pattern, and the model cannot accurately reconstruct it, resulting in a significantly increased reconstruction error. Furthermore, by setting a reasonable error threshold, when the reconstruction error of the time-series data to be detected exceeds the threshold, it is determined to be abnormal.

[0103] For example, in real-world scenarios, the current in line A might exhibit a slow upward or downward trend. This trend could indicate potential risks, such as equipment aging or precursors to partial short circuits. Using the pre-trained anomaly detection model provided in this solution to identify features in the processed time-series data, and / or using the trained LSTM model to identify time-series data, can both achieve accurate identification of potential anomalies.

[0104] S105. Time series data with an abnormality level greater than a preset threshold are identified as potential abnormal data, and an alarm is issued for the potential abnormal data.

[0105] In this step, since the time series data with an anomaly level greater than the preset threshold indicates that the operating mode of the time series data has deviated from the normal operating mode of the substation parameters, it indicates that there is a fault in the equipment or line corresponding to the substation parameters or other equipment or lines associated with them. Therefore, the time series data is identified as potential abnormal data for subsequent alarm and fault location analysis.

[0106] Alarms can be triggered by pushing corresponding alarm information to staff. For example, alarms can be pushed to staff's mobile devices (such as smartphones) or to the monitoring screen in the central control room. Specifically, alarm information may include potential abnormal data and the time when the potential abnormal data occurred.

[0107] The substation anomaly handling method provided in this embodiment acquires time-series data corresponding to multiple substation parameters; preprocesses the time-series data corresponding to each substation parameter to obtain processed time-series data; extracts features from each processed time-series data to obtain the features of each processed time-series data; for each substation parameter, based on the features of the processed time-series data corresponding to the substation parameter, uses a pre-trained anomaly identification model to identify potential anomalies and outputs the anomaly degree of the time-series data; and identifies time-series data with an anomaly degree greater than a preset threshold as potential anomaly data and issues an alarm for that data. This method accurately identifies potential anomaly data in the substation, achieves early warning of substation faults, enables preventive measures to be taken at the nascent stage of the fault, effectively avoids the expansion and deterioration of the fault, improves the efficiency of substation operation and maintenance, reduces power outage time and economic losses, and effectively enhances the safety, stability and reliability of substation operation.

[0108] Furthermore, in one possible implementation, the above embodiments also provide a method for identifying out-of-limit abnormal data, including the following internal steps a to b:

[0109] Step a: For each substation parameter, perform quality optimization on the processed time series data corresponding to that substation parameter to obtain the optimized time series data corresponding to that substation parameter.

[0110] The quality optimization process includes at least one of the following: cleaning, data completion, normalization, dimensionless processing, and noise removal based on data fluctuation.

[0111] It should be understood that the present invention, through quality optimization processing, eliminates noise and defects in data caused by sensor failures, communication interference, etc., providing a reliable and consistent data foundation for subsequent anomaly identification, and avoiding misjudgment or missed judgment due to data quality issues.

[0112] Specifically, the preprocessed time-series data is cleaned to remove outliers and missing values; data is completed using methods such as mean imputation, linear interpolation, and nearest neighbor substitution; data of different dimensions are mapped to a unified interval through normalization; the processed time-series data is divided according to a calculation period set in advance based on the actual fluctuation pattern of the parameters; the mean of the data in each period is calculated, and the fluctuation of the mean of adjacent periods is calculated; targeted filtering is performed on the data in periods with large fluctuations to achieve noise removal based on data fluctuation, resulting in optimized time-series data.

[0113] Step b: In the optimized time-series data corresponding to the substation parameters, data that exceeds the parameter threshold corresponding to the substation parameters are identified as out-of-limit abnormal data, and an alarm is issued for out-of-limit abnormal data.

[0114] In this step, the parameter threshold refers to a threshold predetermined based on the normal variation range of substation parameters. Different substations have different parameter thresholds. If, after quality optimization, the time-series data still contains data exceeding the corresponding parameter threshold, it indicates a fault in the equipment or line corresponding to that parameter, or in other equipment or lines associated with that equipment or line. Therefore, this data is identified as out-of-limit abnormal data.

[0115] This implementation method first performs quality optimization on the processed time-series data, and then performs limit-crossing anomaly identification on the optimized time-series data. This eliminates interference from the acquisition device and communication links, and improves the accuracy and reliability of limit-crossing anomaly identification.

[0116] Figure 2 This is a flowchart illustrating the substation anomaly handling method provided in Embodiment 2 of this application. Based on the above embodiments, this embodiment also provides a fault point analysis method based on anomaly data, specifically including the following:

[0117] S201. Obtain the first abnormal correlation data corresponding to the out-of-limit abnormal data, and the second abnormal correlation data corresponding to the potential abnormal data.

[0118] The first abnormality-related data includes the first protection action signal, fault recording segment, and switch status information during the period when the over-limit abnormality occurred; the second abnormality-related data includes the second protection action signal, fault recording segment, and switch status information during the period when the potential abnormality occurred.

[0119] Specifically, protection action signals (including the first protection action signal and the second protection action signal) refer to the relevant information of the action command signal and feedback signal issued by the relay protection device in the substation, which are used to indicate the protection range, action time and action type of the protection action; fault recording segment refers to the waveform data segment of key electrical and non-electrical quantities over a period of time recorded by the fault recording device; switch status information refers to the real-time status data of switch equipment such as circuit breakers and disconnectors in the substation, which can be obtained through auxiliary contacts or position sensors, mainly including the two states of "closed" and "open", as well as the timestamp of the state transition.

[0120] S202. Determine the fault type of the substation based on the first abnormal correlation data and the second abnormal correlation data.

[0121] It should be understood that determining the type of anomaly is of great significance, as a preliminary assessment can provide a targeted basis for subsequent fault location and analysis.

[0122] In this step, the action signals, fault recording segments, and switch status information during the two periods of abnormal data occurrence can be compared to perform multi-feature cross-validation, preliminarily determine the current fault type, and thus narrow down the subsequent fault point search range.

[0123] Among the fault types are, for example, transformer inter-turn short circuit, single-phase grounding fault, and bushing insulation fault.

[0124] S203. Based on the fault type, a pre-constructed substation power grid topology diagram is generated, and associated nodes in the power grid topology diagram are respectively associated with out-of-limit abnormal data and potential abnormal data. Abnormal intervals or abnormal equipment of the substation are determined in the power grid topology diagram.

[0125] In this pre-constructed power grid topology diagram, the substation connection lines are transformed into a graph consisting of nodes and branches. Generators, transformers, and loads are treated as nodes, and the electrical connections (i.e., lines) between equipment are treated as branches. By determining the hierarchical relationship between nodes in the power grid topology diagram, the path for power transmission can be clearly defined.

[0126] Specifically, the substation topology refers to the connection methods and layout relationships of electrical equipment and lines within a substation, reflecting the physical architecture and electrical connections of the substation. It typically includes the connection forms of key components such as busbars, transformers, circuit breakers, disconnectors, and transmission lines, such as single busbar, double busbar, ring, and radial topologies. This structure determines the power transmission path, distribution method, and isolation range during faults within the substation.

[0127] In a power grid topology diagram, the associated node for out-of-limit abnormal data refers to the node in the power grid topology diagram that is closest to the device / line from which the out-of-limit abnormal data originates; the associated node for potential abnormal data in a power grid topology diagram refers to the node in the power grid topology diagram that is closest to the device / line from which the potential abnormal data originates.

[0128] In this step, it is necessary to initially locate the fault based on the fault type, determine the search direction, and then, based on the associated nodes in the power grid topology diagram that are associated with the out-of-limit abnormal data and potential abnormal data respectively, determine the abnormal interval or abnormal equipment of the substation in the power grid topology diagram.

[0129] Among them, abnormal interval refers to the functional area where the source of the fault exists; abnormal equipment refers to the source equipment where the fault exists.

[0130] In one possible implementation, steps 3.1 to 3.2 can be used:

[0131] Step 3.1: In the pre-constructed power grid topology diagram of the substation, starting from the associated nodes associated with the out-of-limit abnormal data and the associated nodes associated with the potential abnormal data, within the fault association area matching the fault type, based on the electrical connections in the power grid topology diagram and the first protection action information and the second protection action information, a breadth-first search algorithm is used to determine the breakpoint of the substation.

[0132] Among them, the breakpoint is the node that causes the interruption of power transmission in the power grid.

[0133] In this step, the fault association region for fault type matching refers to the range of potentially involved devices and connections defined based on the identified fault types, used to narrow down the search scope. Breadth-first search is a graph search algorithm that, within the defined search range, starts from the starting point, prioritizes visiting all adjacent nodes, and then visits more distant nodes layer by layer.

[0134] In the specific implementation of this scheme, the associated nodes of the out-of-limit abnormal data and the associated nodes of the potential abnormal data are first set as the initial starting point of the breadth-first search algorithm, and the search range of the breadth-first search algorithm is limited according to the fault type. Starting from the starting point, the adjacent nodes are visited layer by layer according to the electrical connection relationship. For each node, its electrical transmission status is checked in combination with the protection action information. If the node is a circuit breaker and the protection action shows tripping, then the node is marked as a breakpoint.

[0135] Step 3.2: Based on the electrical connection of the substation's breakpoint in the power grid topology diagram, obtain the interval associated with the breakpoint as an abnormal interval, or obtain the device associated with the breakpoint as an abnormal device.

[0136] In this step, it is necessary to locate the interval associated with the breakpoint as an abnormal interval based on the upstream and downstream connection relationship of the breakpoint in the power grid topology diagram (such as the transformer upstream of the breakpoint and the bus downstream). Alternatively, it is necessary to locate the equipment associated with the breakpoint as an abnormal device.

[0137] This implementation limits the search range based on the fault type and uses a breadth-first search algorithm to search for breakpoints, reducing the time spent investigating invalid areas and improving the speed of breakpoint location. In addition, the method of determining abnormal equipment or intervals based on breakpoints and their topological relationships provides maintenance personnel with a concrete fault target, effectively avoiding redundant inspections of unrelated equipment, further improving the accuracy of fault location and the response efficiency of fault handling, and providing key technical support for the rapid isolation and recovery of power grid faults.

[0138] S204. Determine the fault point of the substation based on the abnormal intervals and abnormal equipment.

[0139] In this step, after narrowing down the scope to abnormal intervals or abnormal devices, the fault point will be further identified from the abnormal intervals or abnormal devices.

[0140] In one possible implementation, abnormal intervals or abnormal equipment are characterized by traveling wave ranging and / or synchronous phasor measurement techniques, and the fault point of the substation is determined based on the detection results.

[0141] When a fault occurs (such as a short circuit or grounding), a transient traveling wave is generated in the abnormal interval or equipment, propagating along the line / equipment at a speed close to the speed of light, and it is reflected and refracted at the fault point. Therefore, traveling wave ranging technology can detect the arrival time, wavefront polarity, and reflected wave characteristics of the traveling wave, and combine this with the propagation speed to calculate the distance from the fault point of the abnormal interval or abnormal equipment to the measuring device, thereby locating the fault point.

[0142] Meanwhile, synchronous phasor measurement technology can be used to collect voltage and current phasors in abnormal intervals or abnormal equipment in real time through a high-precision synchronous clock. By utilizing the abrupt change characteristics of phasors before and after the fault (such as phase difference and amplitude change rate), combined with the power grid topology and impedance parameters, the fault point can be located.

[0143] This implementation method captures the transient traveling wave propagation characteristics of abnormal objects through traveling wave ranging and / or uses synchronous phasor measurement to perform high-precision analysis of the abrupt change law of electrical phasors. This enables the fault location to be accurate to the specific physical location, greatly improving the spatial accuracy of fault point location and thus improving the efficiency of subsequent maintenance.

[0144] Furthermore, in one possible implementation, the method provided in this embodiment further includes:

[0145] Based on 3D rendering technology, a model of the substation is created; the time-series data of multiple substation parameters and / or fault points are then visualized in the model.

[0146] As a specific example, firstly, a 3D model of equipment such as transformers and circuit breakers is constructed and an image base map is obtained. Then, the model is loaded using the Cesium engine, and the coordinates of the Geographic Information System (GIS) are determined. Finally, real-time data (including abnormal data) is mapped onto the surface of the equipment model through color gradients, thereby realizing a visual display of the operating status and allowing maintenance personnel to intuitively understand the equipment status.

[0147] Cesium is an open-source JavaScript library focused on building high-performance 3D geospatial visualization applications. Based on WebGL technology, it renders global terrain, imagery, 3D models, and vector data in a browser without requiring plugins, supporting full-dimensional spatial analysis from satellite views to underground pipe networks. Cesium supports multiple data formats, such as 3DTiles, KML, and GeoJSON, and possesses powerful large-scale data loading and rendering capabilities. Furthermore, it offers rich interactive features, such as address search, information attribute boxes, and WebVR virtual reality experiences, making it widely used in smart cities, disaster emergency response, environmental monitoring, education, and scientific research.

[0148] Specifically, a color gradient mapping principle can be used, with preset color ranges corresponding to data value ranges. For example, green to yellow to red represents normal, near-abnormal, and abnormal, respectively. Taking transformer temperature abnormality as an example, normal is 50℃-80℃, and above 80℃ is abnormal. Different temperatures correspond to different colors displayed on the model surface, highlighting and labeling the fault location. The same applies to circuit breaker current abnormality. The normal current range is 100A-500A, and above 500A is abnormal. Colors are mapped and labeled according to the rules, facilitating quick problem location and handling by maintenance personnel.

[0149] In addition, after locating the fault point, the fault point can be displayed at the corresponding location on the model to improve the intuitiveness of fault location.

[0150] Optionally, users can combine click event commands, pop-up device details panel commands, zoom commands, and view rotation commands with the visual display of the operating status to show the location of abnormal data. This clearly improves user interactivity.

[0151] The method provided in this implementation improves the intuitiveness of fault display by visualizing the substation in three dimensions, which helps to meet the needs of efficient operation and maintenance in complex scenarios.

[0152] The substation anomaly handling method provided in this embodiment integrates out-of-limit and potential anomaly-related data, and combines fault type and power grid topology to achieve a layer-by-layer focusing approach from data parsing to fault point location. This enables the fault to be located to the precise fault point, thereby achieving the technical effects of improving the accuracy and efficiency of substation fault location, shortening the fault investigation and handling cycle, and enhancing the power grid fault response capability.

[0153] Figure 3 This is a schematic diagram of the substation fault handling device provided in Embodiment 3 of this application, as shown below. Figure 3 As shown, the substation fault handling device 30 provided in this embodiment includes:

[0154] The acquisition module 301 is used to acquire time-series data corresponding to multiple substation parameters of the substation, wherein the different substation parameters are different electrical quantity parameters or non-electrical quantity parameters of the substation.

[0155] The preprocessing module 302 is used to preprocess the time series data corresponding to each substation parameter to obtain the processed time series data.

[0156] The feature extraction module 303 is used to extract features from each processed time series data to obtain the features of each processed time series data.

[0157] The potential anomaly identification module 304 is used to identify potential anomalies for each substation parameter based on the characteristics of the processed time series data corresponding to the substation parameter, using a pre-trained anomaly identification model, and output whether the substation parameter is abnormal. The anomaly identification model is obtained in advance by training a machine learning model based on historical substation data. The machine learning model includes the isolated forest model.

[0158] The first determining module 305 is used to determine the time-series data corresponding to the abnormal substation parameters as potential abnormal data.

[0159] The substation anomaly handling device 30 provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0160] Figure 4 This is a schematic diagram of the substation fault handling device provided in Embodiment 4 of this application, as shown below. Figure 4 As shown, based on the above embodiments, the substation fault handling device 30 provided in this embodiment further includes:

[0161] The out-of-limit anomaly identification module 306 is used to perform quality optimization processing on the processed time series data corresponding to each substation parameter to obtain the optimized time series data corresponding to the substation parameter. The quality optimization processing includes at least one of cleaning, data completion processing, normalization processing, dimensionless processing, and noise removal processing based on data fluctuation.

[0162] Data in the optimized time-series data corresponding to the substation parameters that exceed the parameter thresholds corresponding to the substation parameters are identified as out-of-limit abnormal data, and an alarm is issued for the out-of-limit abnormal data.

[0163] The second determining module 307 includes:

[0164] The acquisition unit is used to acquire first abnormal correlation data corresponding to the over-limit abnormal data and second abnormal correlation data corresponding to the potential abnormal data; the first abnormal correlation data includes a first protection action signal, fault recording segment, and switch status information during the over-limit abnormality period; the second abnormal correlation data includes a second protection action signal, fault recording segment, and switch status information during the potential abnormality period.

[0165] The first determining unit is used to determine the fault type of the substation based on the first abnormal correlation data and the second abnormal correlation data.

[0166] The second determining unit is used to pre-construct a power grid topology map of the substation based on the fault type, wherein the associated nodes in the power grid topology map are respectively associated with the out-of-limit abnormal data and the potential abnormal data, and to determine the abnormal interval or abnormal equipment of the substation in the power grid topology map.

[0167] The third determining unit is used to determine the fault point of the substation based on the abnormal interval or abnormal equipment.

[0168] The display module 308 is used to model the substation based on 3D rendering technology to obtain a model of the substation; and to visualize the time-series data of the multiple substation parameters and / or the fault points in the model.

[0169] In one possible implementation, the features of the processed time-series data in the feature extraction module 303 include at least one of the following: mean, variance, maximum value, minimum value, skewness, kurtosis, dominant frequency amplitude, harmonic distortion rate, outlier percentage, and volatility of the processed time-series data.

[0170] In one possible implementation, the second determining unit is specifically used for:

[0171] In the pre-constructed power grid topology diagram of the substation, starting from the associated nodes associated with the out-of-limit abnormal data and the associated nodes associated with the potential abnormal data, within the fault association area matching the fault type, based on the electrical connections in the power grid topology diagram and the first protection action information and the second protection action information, a breadth-first search algorithm is used to determine the breakpoint of the substation, where the breakpoint is the node that causes the interruption of power transmission in the power grid.

[0172] Based on the electrical connection of the substation's breakpoint in the power grid topology diagram, the interval associated with the breakpoint is obtained as an abnormal interval, or the device associated with the breakpoint is obtained as an abnormal device.

[0173] In one possible implementation, the third determining unit is specifically used for:

[0174] The abnormal intervals or abnormal equipment are characterized by traveling wave ranging and / or synchronous phasor measurement technology, and the fault point of the substation is determined based on the detection results.

[0175] In one possible implementation, the preprocessing in the preprocessing module 302 includes analog-to-digital conversion, filtering, scaling conversion, and recording event sequence processing.

[0176] The substation anomaly handling device 30 provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0177] Figure 5 A schematic diagram of the structure of the computer device provided in this application. Figure 5 As shown, the computer device 40 provided in this embodiment includes at least one processor 401 and a memory 402. Optionally, the device 40 further includes a communication component 403. The processor 401, memory 402, and communication component 403 are connected via a bus 404.

[0178] In a specific implementation, at least one processor 401 executes computer execution instructions stored in memory 402, causing at least one processor 401 to perform the above-described method.

[0179] The specific implementation process of processor 401 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0180] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0181] The memory may include read-only memory and random access memory. The memory may be volatile or non-volatile, or may include both. Non-volatile memory may include read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory may include random access memory (RAM), which serves as an external cache. Many forms of RAM are available by way of example, but not limitation. Examples include Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced Synchronous DRAM (ESDRAM), Sync Link DRAM (SLDRAM), and Direct Rambus RAM (DR RAM).

[0182] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0183] This application also provides a computer program product, including a computer program that, when executed, implements the above-described method.

[0184] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed, implement the above-described method.

[0185] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as SRAM, EEPROM, EPROM, PROM, ROM, magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0186] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside within an ASIC. Alternatively, the processor and the readable storage medium can exist as discrete components in a device.

[0187] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0188] The units described as separate components may or may not be physically separate. The components shown as units 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 units can be selected to achieve the purpose of this embodiment according to actual needs.

[0189] In addition, the functional units 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.

[0190] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an electronic device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0191] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0192] The above embodiments are merely preferred embodiments provided to fully illustrate the present invention, and the scope of protection of the present invention is not limited thereto. Equivalent substitutions or modifications made by those skilled in the art based on the present invention are all within the scope of protection of the present invention.

[0193] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for handling anomalies in a substation, characterized in that, include: Acquire time-series data corresponding to multiple substation parameters of the substation, where different substation parameters are different electrical quantity parameters or non-electrical quantity parameters of the substation; After preprocessing the time series data corresponding to each substation parameter, the processed time series data is obtained. Feature extraction is performed on each processed time series data to obtain the features of each processed time series data; For each substation parameter, based on the characteristics of the processed time series data corresponding to the substation parameter, a pre-trained anomaly identification model is used to identify potential anomalies and output the degree of anomaly of the time series data. The anomaly identification model is obtained in advance by training a machine learning model based on historical substation data. The machine learning model includes the isolated forest model. Time series data with an anomaly level greater than a preset threshold are identified as potentially abnormal data, and an alarm is issued for the potentially abnormal data.

2. The method according to claim 1, characterized in that, The characteristics of the processed time series data include at least one of the following: mean, variance, maximum value, minimum value, skewness, kurtosis, dominant frequency amplitude, harmonic distortion rate, outlier percentage, and volatility.

3. The method according to claim 1 or 2, characterized in that, The method further includes: For each substation parameter, the processed time-series data corresponding to the substation parameter is subjected to quality optimization processing to obtain the optimized time-series data corresponding to the substation parameter. The quality optimization processing includes at least one of cleaning, data completion processing, normalization processing, dimensionless processing, and noise removal processing based on data fluctuation. Data in the optimized time-series data corresponding to the substation parameters that exceed the parameter thresholds corresponding to the substation parameters are identified as out-of-limit abnormal data, and an alarm is issued for the out-of-limit abnormal data.

4. The method according to claim 3, characterized in that, The method further includes: Obtain the first abnormal correlation data corresponding to the over-limit abnormal data, and the second abnormal correlation data corresponding to the potential abnormal data; the first abnormal correlation data includes the first protection action signal, fault recording segment, and switch status information during the time period of the over-limit abnormality; the second abnormal correlation data includes the second protection action signal, fault recording segment, and switch status information during the time period of the potential abnormality. Based on the first abnormal correlation data and the second abnormal correlation data, the fault type of the substation is determined; Based on the fault type, a pre-constructed power grid topology diagram of the substation is constructed, and the associated nodes in the power grid topology diagram are respectively associated with the out-of-limit abnormal data and the potential abnormal data. The abnormal interval or abnormal equipment of the substation is determined in the power grid topology diagram. The fault point of the substation is determined based on the abnormal interval or abnormal equipment.

5. The method according to claim 4, characterized in that, The pre-constructed power grid topology diagram of the substation based on the fault type, wherein the associated nodes in the power grid topology diagram are respectively associated with the out-of-limit abnormal data and the potential abnormal data, and the abnormal intervals or abnormal equipment of the substation are determined in the power grid topology diagram, including: In the pre-constructed power grid topology diagram of the substation, starting from the associated nodes associated with the out-of-limit abnormal data and the associated nodes associated with the potential abnormal data, within the fault association area matching the fault type, based on the electrical connections in the power grid topology diagram and the first protection action information and the second protection action information, a breadth-first search algorithm is used to determine the breakpoint of the substation, where the breakpoint is the node that causes the interruption of power transmission in the power grid. Based on the electrical connection of the substation's breakpoint in the power grid topology diagram, the interval associated with the breakpoint is obtained as an abnormal interval, or the device associated with the breakpoint is obtained as an abnormal device.

6. The method according to claim 4, characterized in that, The step of determining the fault point of the substation based on the abnormal interval or abnormal equipment includes: The abnormal intervals or abnormal equipment are characterized by traveling wave ranging and / or synchronous phasor measurement technology, and the fault point of the substation is determined based on the detection results.

7. The method according to claim 4, characterized in that, The method further includes: The substation is modeled using 3D rendering technology to obtain a model of the substation. The time-series data of the multiple substation parameters and / or the fault points are visualized in the model.

8. The method according to claim 1 or 2, characterized in that, The preprocessing includes analog-to-digital conversion, filtering, scaling, and recording the order of events.

9. An anomaly handling device for a substation, characterized in that, include: The acquisition module is used to acquire time-series data corresponding to multiple substation parameters of the substation, where different substation parameters are different electrical quantity parameters or non-electrical quantity parameters of the substation. The preprocessing module is used to preprocess the time series data corresponding to each substation parameter to obtain the processed time series data. The feature extraction module is used to extract features from each processed time series data to obtain the features of each processed time series data. The potential anomaly identification module is used to identify potential anomalies for each substation parameter based on the characteristics of the processed time-series data corresponding to the substation parameter, using a pre-trained anomaly identification model, and output whether the substation parameter is abnormal. The anomaly identification model is obtained in advance by training a machine learning model based on historical substation data, and the machine learning model includes the isolated forest model. The first determination module is used to identify the time-series data corresponding to the abnormal substation parameters as potential abnormal data.

10. A computer device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-8.

Images