Converter station seismic state monitoring method and device based on clustering neural network

Abstract

The invention discloses a method and device for monitoring the seismic state of a converter station based on a clustering neural network, and relates to the field of seismic state monitoring. The method comprises the following steps: in response to operation state monitoring operation of a user on different types of power equipment in a converter station in an earthquake environment, acquiring parameter information corresponding to each power equipment in an early stage, a middle stage and a later stage of an earthquake; performing time sequence analysis on the parameter information through a multi-scale damage characterization system, and obtaining corresponding time sequence data of the power equipment in the early stage, the middle stage and the later stage of the earthquake; performing feature extraction operation on the time sequence data according to a preset mode, and obtaining parameter change trend information in the time sequence data according to the feature extraction operation; according to the parameter change trend information, the parameter information corresponding to the early stage of earthquake occurrence, the middle stage of earthquake occurrence and the later stage of earthquake occurrence, and according to a group topological graph neural network, obtaining state data corresponding to different types of power equipment in the earthquake environment; the method comprises the following steps: calculating a cluster centroid distance corresponding to each power device according to a clustering statistical neural network by inputting state data; and according to the cluster centroid distance, calculating a comprehensive state evaluation result corresponding to each power device in the converter station through weighting. According to the technical scheme of the invention, the problems of insufficient feature decoupling, coupling effect deficiency, inaccurate abnormal operation state evaluation and the like of the power equipment in a complex earthquake environment through a traditional monitoring method are solved.

Description

Technical Field

[0001] The present application relates to the field of earthquake status monitoring, and in particular to a method and device for monitoring the earthquake status of a converter station based on a clustering neural network. Background Art

[0002] Converter stations, as key facilities in power systems, are responsible for transmitting high-voltage direct current (HVDC) electricity, ensuring the stable operation of the power grid. However, natural disasters such as earthquakes can damage or malfunction equipment within converter stations, impacting the security of power supply.

[0003] Equipment operation is typically monitored through manual inspections or local sensors, relying on the detection of single parameters such as vibration and temperature. However, due to the limitations of monitoring methods, these solutions are unable to fully capture multi-parameter information, and their processing methods are relatively traditional. Consequently, they suffer from slow response speeds and low accuracy. Furthermore, when multiple different types of power equipment are present in a converter station, monitoring and analysis not only takes a long time, but also fails to accurately determine the abnormal operating status of each type of power equipment. This leads to inaccurate assessments of abnormal operation of power equipment in complex environments.

[0004] Therefore, there is an urgent need for a method and device for monitoring the earthquake status of a converter station based on a clustering neural network. Summary of the Invention

[0005] The present application provides a method and device for monitoring the earthquake status of a converter station based on a clustering neural network, which solves the problems of insufficient decoupling of characteristics of power equipment in complex earthquake environments, lack of coupling effect, and inaccurate assessment of abnormal operating status in traditional monitoring methods.

[0006] In a first aspect of the present application, a method for monitoring the earthquake status of a converter station based on a clustering neural network is provided, the method comprising: in response to a user's operation of monitoring the operating status of different types of power equipment in a converter station under an earthquake environment, obtaining parameter information corresponding to each power equipment in the early stage, middle stage and late stage of an earthquake, respectively, the parameter information including vibration information, strain information, acceleration information and temperature information; performing time series analysis on the parameter information through a multi-scale damage characterization system, and obtaining time series data corresponding to the power equipment in the early stage, middle stage and late stage of an earthquake; performing feature extraction operation on the time series data in a preset manner, and performing feature extraction operation on the time series data according to the ... A feature extraction operation is performed to obtain parameter change trend information in the time series data, which includes vibration change trend information, strain change trend information, acceleration change trend information, and temperature change trend information; based on the parameter change trend information, the parameter information corresponding to the early stage, middle stage, and late stage of the earthquake, and based on the group topology neural network, the status data corresponding to different types of power equipment under the earthquake environment are obtained; by inputting the status data, the cluster centroid distance corresponding to each power equipment is calculated according to the cluster statistical neural network; based on the cluster centroid distance, the comprehensive status evaluation result corresponding to each power equipment in the converter station is calculated through weighted calculation.

[0007] Optionally, parameter information corresponding to each power equipment before the earthquake, in the middle and in the late stage of the earthquake is obtained, specifically including: deploying a sensor array along the converter station through a preset fiber optic sensing technology, the preset fiber optic sensing technology including distributed fiber Bragg grating technology and Brillouin optical time domain reflection technology; sensitizing optical cable sensing arrays around different types of power equipment through a preset fiber optic interferometry measurement technology, the preset fiber optic interferometry measurement technology including Michelson interferometry technology and Mach-Zehnder interferometry technology; obtaining parameter information corresponding to each power equipment before the earthquake, in the middle and in the late stage of the earthquake according to the sensor array and the sensitized optical cable sensing array.

[0008] Optionally, before performing time series analysis on parameter information through a multi-scale damage characterization system and obtaining the corresponding time series data of power equipment in the early, middle and late stages of an earthquake, a multi-scale damage characterization system needs to be constructed, specifically including: establishing a cross-scale damage feature encoding model of vibration spectrum-strain gradient-temperature field, and realizing time series evolution modeling of equipment damage status through gated cyclic units.

[0009] Optionally, a time series analysis is performed on parameter information through a multi-scale damage characterization system, and corresponding time series data of the power equipment in the early stage, middle stage and late stage of the earthquake are obtained, specifically including: performing a time series reconstruction operation on the parameter information, and the time series reconstruction operation is used to perform data segmentation on the parameter information according to different time windows; normalizing the parameter information after the time series reconstruction operation to reduce noise interference and enhance the comparability of the parameter information; based on the time step strategy, performing a time series analysis on the normalized parameter information, and obtaining corresponding time series data of the power equipment in the early stage, middle stage and late stage of the earthquake.

[0010] Optionally, status data corresponding to different types of power equipment under earthquake environment are obtained according to the group topology graph neural network, specifically including: obtaining the layout structure of different types of power equipment in the converter station, and encoding the different types of power equipment into graph structure nodes in the group topology graph neural network according to the layout structure; according to the group topology graph neural network, using the graph attention network to model the seismic wave propagation path to capture the vibration coupling effect between different types of power equipment; according to the vibration coupling effect, obtaining status data corresponding to different types of power equipment under earthquake environment.

[0011] Optionally, by inputting status data, the cluster centroid distance corresponding to each power device is calculated according to the cluster statistical neural network, specifically including: based on the clustering algorithm, dividing the status data of each power device into several clusters, each cluster representing an abnormal state of the device; calculating the centroid value corresponding to each cluster, the centroid value is the value corresponding to the optimal center point of the status data in each cluster; based on the centroid value, calculating the cluster centroid distance corresponding to each power device.

[0012] Optionally, calculating the cluster centroid distance corresponding to each power device based on the centroid value specifically includes: calculating the cluster centroid distance corresponding to each power device based on the centroid value using the following formula:

[0013]

[0014] Among them, d i,c is the distance between the centroid of the ith power equipment and the cth cluster, x i,j is the state data of the i-th power equipment in the j-th feature dimension, ω j is the weight coefficient corresponding to the jth feature dimension, n represents the number of feature dimensions, c j is the centroid value of the cth cluster on the jth feature dimension, σ j represents the standard deviation corresponding to the jth feature dimension, λ is the normalization coefficient, m represents the number of abnormal feature information in the parameter change trend information, represents the reference value corresponding to the kth abnormal feature information, and p is the sensitivity adjustment parameter.

[0015] Optionally, based on the cluster centroid distance, the comprehensive status evaluation results corresponding to each power equipment in the converter station are calculated by weighted calculation, specifically including: judging whether the cluster centroid distance is greater than a preset abnormal threshold; if the cluster centroid distance is greater than the preset abnormal threshold, confirming that the comprehensive status evaluation result corresponding to the power equipment is an abnormal operating state; if the cluster centroid distance is less than or equal to the preset abnormal threshold, confirming that the comprehensive status evaluation result corresponding to the power equipment is a normal operating state.

[0016] Optionally, if the cluster centroid distance is greater than a preset abnormal threshold, then after confirming that the comprehensive status assessment result corresponding to the power equipment is an abnormal operating state, the method also includes: obtaining the abnormality type corresponding to the power equipment based on the cluster centroid distance; obtaining the equipment maintenance plan corresponding to the abnormality type in a preset abnormality handling plan database, the preset abnormality handling plan database is used to store the correspondence between the abnormality type and the equipment maintenance plan; and sending the equipment maintenance plan to the user's terminal device.

[0017] In a second aspect of the present application, a converter station earthquake status monitoring device based on a clustering neural network is provided, the device comprising an acquisition module and a processing module, wherein:

[0018] An acquisition module is used to respond to the user's operation of monitoring the operating status of different types of power equipment in the converter station under an earthquake environment, and obtain the parameter information corresponding to each power equipment in the early stage, middle stage and late stage of the earthquake, respectively. The parameter information includes vibration information, strain information, acceleration information and temperature information; perform time series analysis on the parameter information through a multi-scale damage characterization system, and obtain the corresponding time series data of the power equipment in the early stage, middle stage and late stage of the earthquake; perform feature extraction operations on the time series data in a preset manner, and obtain parameter change trend information in the time series data based on the feature extraction operations, and the parameter change trend information includes vibration change trend information, strain change trend information, acceleration change trend information and temperature change trend information; according to the parameter change trend information, the parameter information corresponding to the early stage, middle stage and late stage of the earthquake, and the group topology graph neural network, obtain the status data corresponding to different types of power equipment under the earthquake environment.

[0019] The processing module is used to calculate the cluster centroid distance corresponding to each power device according to the cluster statistical neural network by inputting the status data; and calculate the comprehensive status evaluation result corresponding to each power device in the converter station through weighted calculation based on the cluster centroid distance.

[0020] In the third aspect of the present application, an electronic device is provided, including a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device performs any of the methods described above.

[0021] In a fourth aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and the computer program is executed by a processor to perform any of the above methods.

[0022] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0023] 1. Obtain the parameter information corresponding to each power equipment in each earthquake stage; obtain the time series data corresponding to the power equipment in each earthquake stage; obtain the parameter change trend information in the time series data based on the feature extraction operation; obtain the status data corresponding to different types of power equipment under the earthquake environment based on the parameter change trend information and the parameter information corresponding to each earthquake stage; calculate the cluster centroid distance corresponding to each power equipment based on the cluster statistical neural network by inputting the status data; based on the cluster centroid distance, calculate the comprehensive status evaluation result corresponding to each power equipment in the converter station through weighted calculation, thereby improving the abnormal status recognition accuracy of the power equipment in the complex environment of the converter station by effectively integrating multi-parameter information, overcoming the limitations of the traditional monitoring method of slow response speed, single monitoring means, and inaccurate abnormal evaluation, and improving the intelligence and reliability of the power equipment operation status monitoring under earthquake environment.

[0024] 2. By using preset fiber optic sensing technology, sensor arrays are deployed along the converter station, and by using preset fiber optic interferometry measurement technology, optical cable sensing arrays are enhanced around different types of power equipment. Based on the sensor arrays and the enhanced optical cable sensing arrays, the parameter information corresponding to each power equipment in the early, middle and late stages of an earthquake can be obtained, thereby achieving high-precision, real-time monitoring of different types of power equipment in the converter station and enhancing the ability to obtain power equipment status data in an earthquake environment.

[0025] 3. Based on the clustering algorithm, the status data of each power equipment is divided into several clusters, each cluster represents an abnormal state of the equipment; the centroid value corresponding to each cluster is calculated, and the centroid value is the value corresponding to the optimal center point of the status data in each cluster; based on the centroid value, the cluster centroid distance corresponding to each power equipment is calculated, thereby realizing the refined classification and abnormality identification of the operating status of the power equipment to adapt to different equipment types and their complex operating characteristics, improve the reliability of equipment status assessment in complex seismic environments, and provide key support for subsequent comprehensive status assessment and maintenance decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 1 is a flow chart of a method for monitoring the earthquake status of a converter station based on a clustering neural network provided in an embodiment of the present application;

[0027] Figure 2 1 is a module diagram of a converter station earthquake status monitoring device based on a clustering neural network provided in an embodiment of the present application;

[0028] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0029] Explanation of the reference numerals: 21, acquisition module; 22, processing module; 301, processor; 302, communication bus; 303, user interface; 304, network interface; 305, memory. DETAILED DESCRIPTION

[0030] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0031] The terms used in the following examples of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification of the present application, the singular expressions "a", "an", "said", "above", "the", and "this" are intended to include plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in the present application refers to and includes any or all possible combinations of one or more of the listed items.

[0032] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood to imply or suggest relative importance or implicitly indicate the number of the technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of this application, unless otherwise specified, "plurality" means two or more.

[0033] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0034] Please refer to Figure 1, which shows a flow chart of a method for monitoring the earthquake status of a converter station based on a clustering neural network provided in an embodiment of the present application. The method is applied to a server. The flow chart mainly includes the following steps: step S101 to step S105.

[0035] Step S101 , in response to a user's operation of monitoring the operating status of different types of power equipment in a converter station under an earthquake environment, obtain parameter information corresponding to each power equipment in the early, middle and late stages of an earthquake.

[0036] Specifically, in response to the user's operation of monitoring the operating status of different types of power equipment in the converter station under an earthquake environment, the corresponding parameter information of each power equipment in the early stage, middle stage and late stage of the earthquake is obtained. Specifically, by deploying a distributed optical fiber sensing system and a sensitivity-enhancing optical cable sensing array in the converter station, various parameter data such as vibration information, strain information, acceleration information, temperature information, etc. of various types of power equipment are monitored in real time. This parameter information helps to comprehensively evaluate the status changes of the equipment in different earthquake stages and provides basic data for subsequent equipment health monitoring and anomaly detection. In the embodiments of the present application, the power equipment includes but is not limited to: converter transformers, valve cooling systems, DC field equipment, converter valves, smoothing reactors, etc.

[0037] In a possible embodiment, step S101 also includes: deploying a sensor array along the converter station using preset fiber optic sensing technology, the preset fiber optic sensing technology including distributed fiber Bragg grating technology and Brillouin optical time domain reflectometry technology; utilizing a preset fiber optic interferometry measurement technology to enhance the sensitivity of optical cable sensing arrays around different types of power equipment, the preset fiber optic interferometry measurement technology including Michelson interferometry technology and Mach-Zehnder interferometry technology; and obtaining parameter information corresponding to each power equipment in the early stage, middle stage, and late stage of an earthquake based on the sensor array and the enhanced optical cable sensing array.

[0038] Specifically, distributed fiber Bragg grating technology can measure the strain and temperature changes of equipment by embedding Bragg gratings as sensing elements in optical fibers. For example, through optical fiber sensing systems or enhanced optical cable sensing arrays, it can specifically monitor the coolant flow rate, valve temperature distribution, and thyristor stress of converter valves. This technology can provide high-resolution spatial distribution monitoring and obtain equipment strain information along the entire length of the optical fiber. It is suitable for monitoring the overall stress conditions and local deformation of the equipment. At the same time, Brillouin optical time-domain reflectometry technology can obtain high-precision strain and temperature information by analyzing the scattering characteristics of light waves in optical fibers. It can perform real-time monitoring over long distances and is suitable for long-term monitoring of large-scale power equipment. Through the combination of these technologies, the optical fiber sensing array can perceive the multi-dimensional status of the equipment in real time and provide reliable raw data for subsequent analysis.

[0039] In addition, fiber optic interferometry technology is pre-installed, with sensor arrays of enhanced optical cables deployed around various types of power equipment to further enhance monitoring accuracy and sensitivity. Fiber optic interferometry technologies include Michelson interferometry and Mach-Zehnder interferometry. Michelson interferometry compares the phase difference between two beams of light through the principle of interference and is suitable for measuring minute displacement changes, such as surface vibrations. Mach-Zehnder interferometry effectively measures changes in temperature and strain, and can simultaneously obtain responses from multiple signal sources, offering high interference resistance and accuracy.

[0040] Step S102 , performing time series analysis on parameter information through a multi-scale damage characterization system, and obtaining corresponding time series data of power equipment in the early stage, middle stage and late stage of an earthquake.

[0041] Specifically, the parameter information is analyzed in time series to obtain the corresponding time series data of each power equipment in the early stage, middle stage and late stage of the earthquake. Specifically, first, by collecting the physical parameter information such as vibration, strain, acceleration, temperature, etc. obtained by the power equipment in different earthquake stages (early stage, middle stage and late stage), as well as the electrical parameter information such as DC voltage fluctuation of the flow station, AC bus voltage distortion rate, current imbalance, etc., these data are processed using time series analysis methods (such as fast Fourier transform, filtering and other technologies). Through time series analysis, the parameter change trend at different time points can be extracted, the response characteristics of the equipment during the earthquake can be identified, and the time series data of each stage can be obtained to provide a data basis for subsequent status assessment and anomaly detection.

[0042] In a possible implementation, step S102 further includes: establishing a cross-scale damage feature encoding model of vibration spectrum-strain gradient-temperature field, and implementing temporal evolution modeling of the device damage state through a gated recurrent unit.

[0043] Specifically, based on the frequency domain feature extraction of vibration spectrum signals, the spatiotemporal coupling analysis of strain gradient distribution, and the modeling of temperature field evolution trends, a multi-dimensional damage characterization feature vector with cross-scale fusion expression is constructed; the feature vector is input into the gated recurrent unit (GRU) network, and the dynamic modeling of the potential damage process of the equipment is realized through the memory mechanism and state update mechanism of the damage feature sequence in the historical time window; the attention mechanism is adopted in the modeling process to enhance the response capability to changes in key time segments, and a damage evolution representation with deep time dependence is obtained through multi-layer stacked GRU units; finally, a high-dimensional time series feature tensor representing the equipment state evolution trajectory during the entire earthquake process is output, providing accurate and engineering-judgable basic data support for subsequent feature extraction operations.

[0044] In a possible embodiment, step S102 also includes: performing a time series reconstruction operation on the parameter information, the time series reconstruction operation is used to perform data segmentation on the parameter information according to different time windows; normalizing the parameter information after the time series reconstruction operation to reduce noise interference and enhance the comparability of the parameter information; based on the time step strategy, performing a time series analysis on the normalized parameter information, and obtaining the corresponding time series data of the power equipment in the early stage, middle stage and late stage of the earthquake.

[0045] Specifically, a time series reconstruction operation is performed on the parameter information to segment the time series data into different time windows, ensuring that the dynamic changes of the equipment during the earthquake can be effectively captured. In the time series reconstruction operation, the parameter information (such as vibration, strain, acceleration, temperature, etc.) in the early, middle, and late stages of the earthquake is first divided into multiple small time segments, usually based on preset time windows, including but not limited to every 10 seconds, 30 seconds, or 1 minute. The raw data is segmented to better process and analyze the detailed dynamic response of the equipment at different earthquake stages.

[0046] Next, the parameter information after the time series reconstruction operation is normalized. The purpose of normalization is to convert different types of parameter data (vibration, strain, acceleration, temperature, etc.) into the same standard range, usually normalized to the interval of [0,1] or [-1,1]. The embodiments in this application do not limit the normalized interval and can be selected according to actual needs. This step helps to reduce the scale differences between different parameters, eliminate the interference caused by different data dimensions, enhance the comparability of each parameter in subsequent analysis, and ensure that each parameter has a balanced impact on the final analysis results.

[0047] After normalization is complete, time series analysis is performed on the normalized parameter information based on the time step strategy. The time step strategy determines the data length at each time point in the time series analysis and the step size of the data cutting (for example, an analysis window every 5 seconds or 10 seconds), which helps capture subtle changes in the device status at different time points. Time series analysis methods can use common techniques such as sliding window method, Fourier transform, time domain analysis, etc. to analyze the change patterns of various parameters in different time periods and identify the characteristic performance of the equipment in the early, middle and late stages of the earthquake.

[0048] Step S103: performing a feature extraction operation on the time series data according to a preset method, and obtaining parameter change trend information in the time series data based on the feature extraction operation.

[0049] Specifically, in this step, first, representative important features are extracted from the time series data obtained in step S102. These features can reflect the dynamic response of power equipment in the early, middle and late stages of an earthquake. Feature extraction methods include but are not limited to the following: Time domain feature extraction: By calculating the basic statistical features of the time series data, such as mean, standard deviation, maximum, minimum, skewness, kurtosis, etc., the basic distribution of the data is obtained; Frequency domain feature extraction: Using methods such as Fourier transform (FFT), the time series data is converted into the frequency domain, the intensity and distribution of different frequency components in the data are analyzed, and information such as the vibration frequency and amplitude changes of the equipment are identified; Time-frequency feature extraction: By using methods such as wavelet transform, the local time-frequency features of the time series data are extracted. The above method is very effective for capturing instantaneous changes in signals such as vibration and acceleration, and is particularly suitable for non-stationary signal analysis; Trend analysis features: By using trend detection methods (such as sliding average, linear regression, etc.), the long-term change trend of the equipment in different time periods is extracted. For example, monitoring the long-term change trend of temperature or strain helps to identify whether the equipment is gradually becoming abnormal.

[0050] After obtaining the time series characteristics of each parameter (such as vibration, strain, acceleration, and temperature) through the above feature extraction process, further analysis is performed on the changing trends of each parameter. This trend information reflects the equipment's response during the earthquake and can reveal whether any abnormal changes have occurred in the equipment's operating status. Parameter change trends can be analyzed from the following perspectives: Vibration change trends: By extracting and analyzing the characteristics of vibration signals, changes in the equipment's vibration amplitude and frequency can be identified. During an earthquake, the equipment's vibration response will vary, potentially showing a short-term increase in amplitude or a change in frequency, which in turn indicates an abnormal condition. Strain change trends: Changes in strain in power equipment reflect the stress state of the equipment. Trend analysis of strain data can identify whether the equipment has been subjected to excessive external forces, resulting in deformation, and thus determine whether the equipment is at risk of potential damage. Acceleration change trends: Acceleration signals reflect the shock and vibration experienced by the equipment during the earthquake. Acceleration change trend analysis can understand the dynamic response of the equipment and identify any excessive vibration or irregular patterns. Temperature change trends: Temperature changes in the equipment are an important indicator of its operating condition. During an earthquake, temperature changes in equipment may be caused by external vibrations, load changes, and other factors. Therefore, monitoring temperature change trends can help identify whether the equipment is affected or experiencing abnormal phenomena such as overheating.

[0051] Different parameters can be weighted and integrated based on their changing trends to obtain comprehensive trend information. For example, different trend information can be fused based on the relative importance or weight of each parameter to form a comprehensive equipment status change trend assessment model. This comprehensive assessment not only reflects the changes in each individual parameter but also, through multi-dimensional data fusion, identifies overall changes in equipment status. Ultimately, the parameter change trend information extracted from the time series data is summarized to generate an equipment status change trend report. This report details the equipment's operating status before, during, and after the earthquake, as well as the changing trends of each parameter. This provides important data support and reference for subsequent status assessment, anomaly detection, and fault diagnosis.

[0052] Step S104 , according to the parameter change trend information, the parameter information corresponding to the early stage, the middle stage and the late stage of the earthquake, and the group topology neural network, obtain the status data corresponding to different types of power equipment in the earthquake environment.

[0053] Specifically, based on parameter change trend information and parameter information corresponding to the early, middle, and late stages of an earthquake, the state data of different types of power equipment in an earthquake environment is obtained. This state data includes not only the change trend information of the parameter information but also a snapshot of the equipment's operating status at a specific moment. State data generally includes: instantaneous parameter values (such as the current vibration acceleration, temperature, and current), historical change trends (such as the RMS vibration value over the past 10 seconds and the temperature growth rate over the past hour), abnormal characteristics (such as the magnitude of strain data deviation from normal values), and cluster statistical information (such as the state clustering results of multiple sensors, indicating whether they belong to an abnormal cluster).

[0054] In a possible implementation, step S104 also includes: obtaining the layout structure of the different types of power equipment in the converter station, and encoding the different types of power equipment into graph structure nodes in the group topology graph neural network according to the layout structure; based on the group topology graph neural network, using a graph attention network to model the seismic wave propagation path to capture the vibration coupling effect between the different types of power equipment; based on the vibration coupling effect, obtaining the status data corresponding to the different types of power equipment under the earthquake environment.

[0055] Specifically, based on converter station design drawings and operational layout data, a spatial distribution map of power equipment is constructed, extracting the physical adjacency relationships, functional dependencies, and electrical connection paths between devices. Based on this layout structure, each device node is mapped to a graph structure node in a graph neural network. Category attribute labels and initial feature vectors are assigned to different types of power equipment in the graph structure. Subsequently, based on the edge weights in the group topology graph neural network, which represent the vibration coupling strength between devices, a graph attention mechanism is introduced to weightedly model the dynamic response differences between adjacent devices during different stages of seismic wave propagation, thereby identifying key coupling paths and typical vibration response patterns. To enhance the model's ability to analyze complex vibration propagation paths, a multi-level graph attention module is introduced to capture cross-level vibration propagation effects. A joint spatial-temporal feature enhancement mechanism is then used to dynamically update node states. Furthermore, node-level state representation vectors are calculated by combining the structural response sensitivity, coupling strength spectral distribution, and historical vibration feedback data for each type of power equipment. This data reflects the comprehensive response characteristics of each type of power equipment under the influence of the seismic environment, serving as input for the subsequent cluster statistical neural network.

[0056] Step S105 : By inputting the state data, the cluster centroid distance corresponding to each power device is calculated according to the cluster statistical neural network.

[0057] Specifically, the device status data generated in step S104 (including instantaneous parameter values, historical change trends, abnormal characteristics, clustering statistical information, etc.) is used as input and passed into the clustering statistical neural network. Based on the input data, the clustering statistical neural network groups the device status data to generate multiple clusters. Each cluster represents a group of devices in similar states. The clustering algorithm calculates the distance between each power device and the centroid of its cluster (i.e., the center point of the cluster) to obtain the cluster centroid distance. The cluster centroid distance reflects the degree of deviation between the device status and the cluster center. A larger value indicates a more abnormal device status, while a smaller value indicates a relatively normal device status. Ultimately, the cluster centroid distance calculated by the clustering statistical neural network will serve as one of the bases for device status assessment, providing data support for subsequent device health status judgment and fault diagnosis.

[0058] In a possible implementation, step S105 also includes: based on a clustering algorithm, dividing the status data of each power device into several clusters, each cluster representing an abnormal state of the device; calculating the centroid value corresponding to each cluster, the centroid value being the value corresponding to the optimal center point of the status data in each cluster; and calculating the cluster centroid distance corresponding to each power device based on the centroid value.

[0059] Specifically, the status data of each power device is first clustered using a pre-defined clustering algorithm (such as K-means, DBSCAN, or spectral clustering). The goal of clustering is to group device data with similar status into the same cluster, while assigning device data with different statuses to different clusters. Each cluster represents a device abnormality, including but not limited to: valve cooling system failure, reactor winding fracture, and converter transformer oil tank deformation. The status of devices within a cluster is relatively similar. After clustering is complete, the next step is to calculate the centroid of each cluster. The centroid is the optimal center point of all device status data within the cluster, representing the "average" or "center" value of the device status within that cluster. The centroid is typically calculated as the mean or weighted average of all device status data within the cluster. For example, for multi-dimensional status data (such as vibration, temperature, and acceleration), the centroid is the mean of the device status data for these parameters within the cluster. For each power device, the distance between its status data and the centroid of its cluster is calculated; this distance is the cluster centroid distance. The cluster centroid distance can be calculated by Euclidean distance. The cluster centroid distance corresponding to each power device is calculated based on the centroid value using the following formula:

[0060]

[0061] Among them, d i,c is the distance between the centroid of the ith power equipment and the cth cluster, x i,j is the state data of the i-th power equipment in the j-th feature dimension, ω j is the weight coefficient corresponding to the jth feature dimension, n represents the number of feature dimensions, c j is the centroid value of the cth cluster on the jth feature dimension, σ j represents the standard deviation corresponding to the jth feature dimension, λ is the normalization coefficient, m represents the number of abnormal feature information in the parameter change trend information, represents the reference value corresponding to the kth abnormal feature information, and p is the sensitivity adjustment parameter. The calculated cluster centroid distance reflects the difference between the device status data and the normal operating state (or other abnormal state). If the cluster centroid distance is small, it means that the device status is close to the cluster center, indicating that the device is operating normally; if the cluster centroid distance is large, it means that the device status deviates from the normal range and may be abnormal or faulty. These cluster centroid distances can be used to evaluate the device status and determine whether it is in normal operation or abnormal. Based on the calculated cluster centroid distance, the device status can be further evaluated. When the cluster centroid distance of a device exceeds a certain threshold, it indicates that the device may be in an abnormal state and requires further fault diagnosis or maintenance measures. Different devices may have different abnormality thresholds. Cluster analysis helps to dynamically adjust these thresholds to adapt to the operating conditions and characteristics of different devices.

[0062] Step S106: Based on the cluster centroid distance, a comprehensive status evaluation result corresponding to each power equipment in the converter station is calculated by weighted calculation.

[0063] Specifically, the cluster centroid distance of each power equipment represents the distance between the equipment state and the normal or abnormal state, and determines whether the cluster centroid distance is greater than the preset abnormal threshold; if the cluster centroid distance is greater than the preset abnormal threshold, the comprehensive state evaluation result corresponding to the power equipment is confirmed to be an abnormal operating state; if the cluster centroid distance is less than or equal to the preset abnormal threshold, the comprehensive state evaluation result corresponding to the power equipment is confirmed to be a normal operating state. Generally, the preset abnormal threshold can be set by, but not limited to, the following methods: Based on statistical methods: for example, the cluster centroid distance can be set to be abnormal if it exceeds a certain standard deviation of the mean of the centroid distance within the cluster; Based on historical data: by analyzing the distribution of cluster centroid distances of the equipment during normal operation, a normal range is set, and anything beyond this range is abnormal; Based on machine learning models: using supervised learning or unsupervised learning methods, the model is trained to automatically identify anomalies. The comprehensive state evaluation result is a comprehensive health score for each power equipment obtained by weighting the cluster centroid distance, and the formula can be expressed as:

[0064]

[0065] Among them, S is the comprehensive health score corresponding to all power equipment, q is the total number of types of power equipment, and w i is the weight corresponding to the i-th type of power equipment. The comprehensive evaluation result reflects the overall operating status of the equipment. A higher value indicates that the equipment is in an abnormal state, and a lower value indicates that the equipment is in a good state.

[0066] In a possible implementation, step S106 also includes: obtaining the abnormality type corresponding to the power equipment based on the cluster centroid distance; obtaining the equipment maintenance plan corresponding to the abnormality type in a preset abnormality handling plan database, the preset abnormality handling plan database is used to store the correspondence between the abnormality type and the equipment maintenance plan; and sending the equipment maintenance plan to the user's terminal device.

[0067] Specifically, the abnormality type is judged by different feature values, feature library (multi-scale damage characterization system). In the embodiment of the present application, the method for confirming the abnormality type includes but is not limited to the following methods: Based on cluster label: If the cluster centroid distance of the device exceeds the threshold, the characteristics of the cluster to which it belongs can be checked, and the abnormality type can be determined based on the characteristics of the cluster. For example, some clusters may represent temperature abnormalities, and some clusters represent vibration abnormalities; Based on distance analysis: By calculating the distance between the device status and the centroids of multiple clusters, it can be determined which type of abnormal state the device is closer to. For example, the device status may be close to the "overload" cluster, "overheating" cluster or "vibration abnormality" cluster; Multi-dimensional abnormality analysis: According to the changing trends of multiple parameters (such as temperature, vibration, acceleration, etc.), the abnormality type is judged. For example, increased temperature and increased acceleration may indicate device overload or component wear. The preset abnormality handling solution database is a database that stores the mapping relationship between abnormality types and equipment maintenance solutions. Its purpose is to automatically obtain the most appropriate maintenance solution based on the detected abnormality type. The construction of this database involves the following aspects: First, the abnormality handling solution database requires the introduction of converter station-specific maintenance rules. This rule base is optimized using expert experience. Therefore, all possible abnormality types need to be defined, such as vibration abnormalities (including converter valve vibration abnormalities, converter transformer vibration abnormalities, and smoothing reactor vibration abnormalities); strain abnormalities (including converter valve cooling system piping strain abnormalities, converter transformer casing strain abnormalities, and GIS (gas-insulated switchgear) strain abnormalities); and temperature abnormalities (including converter valve temperature abnormalities, converter transformer oil temperature abnormalities, and busbar temperature abnormalities). Each type of abnormality can also be divided into different severity levels (such as minor, moderate, and severe). For each abnormality type and severity, a corresponding maintenance plan is developed. For example, a minor vibration abnormality might recommend inspecting the connection and performing preventive maintenance; whereas a severe temperature abnormality might require immediate shutdown for inspection and replacement of the faulty component.Establish a mapping relationship between abnormality types and equipment maintenance plans, for example: abnormal vibration -> check vibration sensors, clean equipment, check motors: check the converter valve support structure, confirm the tightening status of the bolts, and re-reinforce or adjust the support structure if necessary; monitor the vibration mode of the converter transformer through a fiber optic sensing system or vibration sensor. If the vibration exceeds the standard, check the winding support structure to confirm whether reinforcement or replacement of buffer materials is required; check the fixing bracket of the smoothing reactor, adjust the fixing bolts and add damping measures to reduce the impact of vibration, etc.; abnormal temperature -> check the cooling system, check the temperature control equipment, clean the radiator: monitor the operating status of the converter valve cooling system, including the coolant flow rate, cooling fan speed, etc. If an abnormality is found, check the coolant replenishment and whether the cooling system pipes are blocked. If necessary, replace the coolant or clean the cooling channel; Abnormal oil temperature is analyzed to check the oil circulation system for proper function. Oil samples are taken, if necessary, to identify signs of aging or impurity deposits. Thermal imaging is used to monitor busbar temperature distribution. If localized overheating is detected, loose joints should be checked and re-tightened or replaced. Abnormal strain -> structural inspection, repair or replacement of damaged components: Fiber Bragg Grating sensors are used to monitor the strain distribution of converter valve cooling ducts. If localized strain exceeds the specified limit, loose connections are checked and, if necessary, the damaged ducts or supporting structures are replaced. External stress testing is performed on the converter transformer tank. If significant deformation is detected, adjustments to the external support structure should be considered, and the integrity of the tank welds and seals should be checked. Abnormal strain in GIS equipment should be checked for seal integrity and SF6 gas leak testing. If necessary, localized repair welding or seal replacement should be performed. In the pre-set abnormality handling plan database, abnormality types and equipment maintenance plans are established based on pre-defined rules. For example, if the cluster centroid distance is greater than the preset anomaly threshold and the anomaly type is "vibration anomaly," the corresponding equipment maintenance plan might be "check the vibration sensor and equipment connection parts." If the cluster centroid distance is greater than the preset anomaly threshold and the anomaly type is "temperature anomaly," the corresponding equipment maintenance plan might be "check the cooling system and temperature control device." Once the equipment anomaly type is determined and the corresponding maintenance plan is obtained, the system will send the maintenance plan to the user's terminal device (such as the operator's mobile phone, monitoring system, or maintenance personnel's work terminal) through the appropriate communication channel. This allows users to obtain timely equipment fault information and the necessary maintenance measures, allowing them to respond quickly and take action to ensure the continued stable operation of the equipment.

[0068] The present application adopts the above method to obtain the parameter information corresponding to each power equipment in each earthquake stage; obtain the time series data corresponding to the power equipment in each earthquake stage; obtain the parameter change trend information in the time series data based on the feature extraction operation; obtain the status data corresponding to different types of power equipment under the earthquake environment based on the parameter change trend information and the parameter information corresponding to each earthquake stage; calculate the cluster centroid distance corresponding to each power equipment based on the cluster statistical neural network by inputting the status data; and calculate the comprehensive status evaluation result corresponding to each power equipment in the converter station through weighted calculation based on the cluster centroid distance, thereby improving the abnormal status identification accuracy of the power equipment in the complex environment of the converter station by effectively fusing multi-parameter information, overcoming the limitations of traditional monitoring methods such as slow response speed, single monitoring means, and inaccurate abnormal evaluation, and improving the intelligence and reliability of power equipment operation status monitoring under earthquake environment.

[0069] Please refer to Figure 2 , which shows a module schematic diagram of a converter station earthquake status monitoring device based on a clustering neural network provided by an embodiment of the present application. The device is a server, and the server includes an acquisition module 21 and a processing module 22, wherein,

[0070] The acquisition module 21 is used to respond to the user's operation of monitoring the operating status of different types of power equipment in the converter station under an earthquake environment, and obtain the parameter information corresponding to each power equipment in the early stage, middle stage and late stage of the earthquake, respectively. The parameter information includes vibration information, strain information, acceleration information and temperature information; perform time series analysis on the parameter information through a multi-scale damage characterization system, and obtain the corresponding time series data of the power equipment in the early stage, middle stage and late stage of the earthquake; perform feature extraction operation on the time series data according to a preset method, and obtain parameter change trend information in the time series data based on the feature extraction operation, and the parameter change trend information includes vibration change trend information, strain change trend information, acceleration change trend information and temperature change trend information; according to the parameter change trend information, the parameter information corresponding to the early stage, middle stage and late stage of the earthquake, and the group topology graph neural network, obtain the status data corresponding to different types of power equipment under the earthquake environment.

[0071] The processing module 22 is used to calculate the cluster centroid distance corresponding to each of the power equipment according to the cluster statistical neural network by inputting the status data; and calculate the comprehensive status evaluation result corresponding to each of the power equipment in the converter station according to the cluster centroid distance through weighted calculation.

[0072] In one possible embodiment, the acquisition module 21 is used to obtain parameter information corresponding to each power equipment in the early stage, middle stage and late stage of an earthquake, specifically including: through preset fiber optic sensing technology, a sensor array is deployed along the converter station, and the preset fiber optic sensing technology includes distributed fiber Bragg grating technology and Brillouin optical time domain reflection technology; through preset fiber optic interferometry measurement technology, a sensitized optical cable sensing array is set up around different types of power equipment, and the preset fiber optic interferometry measurement technology includes Michelson interferometry technology and Mach-Zehnder interferometry technology; according to the sensor array and the sensitized optical cable sensing array, the parameter information corresponding to each power equipment in the early stage, middle stage and late stage of an earthquake is obtained.

[0073] In one possible implementation, the acquisition module 21 is used to construct a multi-scale damage characterization system before performing time series analysis on parameter information through a multi-scale damage characterization system and obtaining the corresponding time series data of the power equipment before, during, and after an earthquake. Specifically, the system includes: establishing a cross-scale damage feature encoding model of vibration spectrum-strain gradient-temperature field, and realizing time series evolution modeling of the equipment damage state through a gated cyclic unit.

[0074] In one possible embodiment, the acquisition module 21 is used to perform time series analysis on parameter information through a multi-scale damage characterization system, and obtain corresponding time series data of the power equipment in the early stage, middle stage and late stage of the earthquake, specifically including: performing a time series reconstruction operation on the parameter information, the time series reconstruction operation is used to perform data segmentation on the parameter information according to different time windows; normalizing the parameter information after the time series reconstruction operation to reduce noise interference and enhance the comparability of the parameter information; based on the time step strategy, performing time series analysis on the normalized parameter information, and obtaining corresponding time series data of the power equipment in the early stage, middle stage and late stage of the earthquake.

[0075] In one possible embodiment, the processing module 22 is used to obtain status data corresponding to different types of power equipment under an earthquake environment based on a group topology graph neural network, specifically including: obtaining the layout structure of different types of power equipment in the converter station, and encoding the different types of power equipment into graph structure nodes in the group topology graph neural network according to the layout structure; based on the group topology graph neural network, using a graph attention network to model the seismic wave propagation path to capture the vibration coupling effect between different types of power equipment; based on the vibration coupling effect, obtaining status data corresponding to different types of power equipment under an earthquake environment.

[0076] In one possible embodiment, the processing module 22 is used to calculate the cluster centroid distance corresponding to each power device according to the cluster statistical neural network by inputting the status data, specifically including: based on the clustering algorithm, dividing the status data of each power device into several clusters, each cluster representing an abnormal state of the device; calculating the centroid value corresponding to each cluster, the centroid value is the value corresponding to the optimal center point of the status data in each cluster; and calculating the cluster centroid distance corresponding to each power device based on the centroid value.

[0077] In a possible implementation, the processing module 22 is configured to calculate the cluster centroid distance corresponding to each power device based on the centroid value, specifically including: calculating the cluster centroid distance corresponding to each power device based on the centroid value using the following formula:

[0078]

[0079] Among them, d i,c is the distance between the centroid of the ith power equipment and the cth cluster, x i,j is the state data of the i-th power equipment in the j-th feature dimension, ω j is the weight coefficient corresponding to the jth feature dimension, n represents the number of feature dimensions, c j is the centroid value of the cth cluster on the jth feature dimension, σ j represents the standard deviation corresponding to the jth feature dimension, λ is the normalization coefficient, m represents the number of abnormal feature information in the parameter change trend information, represents the reference value corresponding to the kth abnormal feature information, and p is the sensitivity adjustment parameter.

[0080] In one possible embodiment, the processing module 22 is used to calculate the comprehensive status evaluation results corresponding to each power equipment in the converter station through weighted calculation based on the cluster centroid distance, specifically including: judging whether the cluster centroid distance is greater than a preset abnormal threshold; if the cluster centroid distance is greater than the preset abnormal threshold, confirming that the comprehensive status evaluation result corresponding to the power equipment is an abnormal operating state; if the cluster centroid distance is less than or equal to the preset abnormal threshold, confirming that the comprehensive status evaluation result corresponding to the power equipment is a normal operating state.

[0081] In one possible embodiment, the processing module 22 is used to confirm that the comprehensive status assessment result corresponding to the power equipment is an abnormal operating state if the cluster centroid distance is greater than a preset abnormal threshold. The method also includes: obtaining the abnormality type corresponding to the power equipment based on the cluster centroid distance; obtaining the equipment maintenance plan corresponding to the abnormality type in the preset abnormality handling plan database, the preset abnormality handling plan database is used to store the correspondence between the abnormality type and the equipment maintenance plan; and sending the equipment maintenance plan to the user's terminal device.

[0082] It should be noted that the above embodiments provide devices that implement their functions using only the division of the above functional modules as examples. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiments are based on the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0083] This application also provides an electronic device. Figure 3 , Figure 3 3 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. The electronic device may include: at least one processor 301, at least one communication bus 302, a user interface 303, at least one network interface 304, and a memory 305.

[0084] The communication bus 302 is used to implement the connection and communication between these components.

[0085] The user interface 303 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 303 may also include a standard wired interface and a wireless interface.

[0086] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0087] The processor 301 may include one or more processing cores. The processor 301 utilizes various interfaces and circuits to connect various components within the server. By running or executing instructions, programs, code sets, or instruction sets stored in the memory 305, and by accessing data stored in the memory 305, the processor 301 performs various server functions and processes data. Optionally, the processor 301 may be implemented using at least one hardware form selected from the group consisting of a digital signal processing (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA). The processor 301 may integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU primarily processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; and the modem is responsible for handling wireless communications. It is understood that the modem may not be integrated into the processor 301 and may be implemented separately on a single chip.

[0088] Among them, the memory 305 may include a random access memory (RAM) or a read-only memory (Read-Only Memory). Optionally, the memory 305 includes a non-transitory computer-readable storage medium. The memory 305 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 305 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 305 may also be optionally at least one storage device located away from the aforementioned processor 301. Refer to Figure 3 The memory 305 as a computer storage medium may include an operating system, a network communication module, a user interface module, and an application for converter station earthquake status monitoring based on clustering neural network.

[0089] exist Figure 3In the electronic device shown, the user interface 303 is mainly used to provide an input interface for the user and obtain the data input by the user; and the processor 301 can be used to call the application for converter station earthquake status monitoring based on clustering neural network stored in the memory 305. When executed by one or more processors 301, the electronic device executes one or more methods as described in the above embodiments. It should be noted that for the aforementioned method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should know that this application is not limited by the order of the actions described, because according to this application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for this application.

[0090] The present application also provides a computer-readable storage medium storing instructions, which, when executed by one or more processors, enable an electronic device to execute one or more of the methods described in the above embodiments.

[0091] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0092] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are merely schematic, such as the division of units, which is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interface, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0093] Units described as separate components may or may not be physically separate, and 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 these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0094] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0095] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a memory and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned memory includes various media that can store program codes, such as USB flash drives, mobile hard drives, magnetic disks or optical disks.

[0096] The foregoing description is merely an exemplary embodiment of the present disclosure and does not limit the scope of the present disclosure. In other words, any equivalent variations and modifications made based on the teachings of the present disclosure are still within the scope of the present disclosure. Those skilled in the art will readily conceive of other embodiments of the present disclosure after considering the disclosure and practical experience.

[0097] This application is intended to cover any modifications, uses or adaptations disclosed in this application, which follow the general principles disclosed in this application and include common knowledge or customary technical means in the technical field not disclosed in this application.

Claims

1. A method for monitoring the earthquake status of a converter station based on clustering neural network, characterized in that: The method comprises: In response to a user's operation to monitor the operating status of different types of power equipment in a converter station under an earthquake environment, parameter information corresponding to each of the power equipment in the early, middle, and late stages of an earthquake is obtained, the parameter information including vibration information, strain information, acceleration information, and temperature information; Performing time series analysis on the parameter information through a multi-scale damage characterization system, and obtaining corresponding time series data of the power equipment in the early stage, the middle stage and the late stage of the earthquake; Performing a feature extraction operation on the time series data in a preset manner, and obtaining parameter change trend information in the time series data based on the feature extraction operation, wherein the parameter change trend information includes vibration change trend information, strain change trend information, acceleration change trend information, and temperature change trend information; According to the parameter change trend information, the parameter information corresponding to the early stage of the earthquake, the middle stage of the earthquake, and the late stage of the earthquake, and according to the group topology neural network, the status data corresponding to the different types of power equipment in the earthquake environment are obtained; By inputting the state data, calculating the cluster centroid distance corresponding to each of the power devices according to the cluster statistical neural network; According to the cluster centroid distance, a comprehensive status evaluation result corresponding to each of the power equipment in the converter station is calculated by weighted calculation.

2. The method according to claim 1, characterized in that The obtaining of parameter information corresponding to each of the power equipment in the early stage of an earthquake, the middle stage of an earthquake, and the late stage of an earthquake specifically includes: Deploying a sensor array along the converter station using a preset fiber optic sensing technology, wherein the preset fiber optic sensing technology includes a distributed fiber Bragg grating technology and a Brillouin optical time domain reflectometry technology; By using a preset fiber optic interferometry measurement technology, a sensitive optical cable sensing array is enhanced around different types of power equipment, wherein the preset fiber optic interferometry measurement technology includes Michelson interferometry and Mach-Zehnder interferometry; According to the sensing array and the enhanced optical cable sensing array, the parameter information corresponding to each of the power equipment in the early stage of the earthquake, the middle stage of the earthquake, and the late stage of the earthquake is obtained.

3. The method according to claim 1, characterized in that Before performing time series analysis on the parameter information through the multi-scale damage characterization system and obtaining the corresponding time series data of the power equipment in the early stage, the middle stage and the late stage of the earthquake, it is necessary to construct the multi-scale damage characterization system, which specifically includes: A cross-scale damage feature encoding model of vibration spectrum-strain gradient-temperature field is established, and the temporal evolution modeling of equipment damage state is realized through gated recurrent units.

4. The method according to claim 1, wherein The performing time series analysis on the parameter information by using a multi-scale damage characterization system and obtaining corresponding time series data of the power equipment in the early stage, the middle stage and the late stage of the earthquake specifically includes: Performing a time series reconstruction operation on the parameter information, wherein the time series reconstruction operation is used to perform data segmentation on the parameter information according to different time windows; performing normalization processing on the parameter information after the time series reconstruction operation to reduce noise interference and enhance the comparability of the parameter information; Based on the time step strategy, the parameter information after the normalization process is subjected to time series analysis, and the time series data corresponding to the power equipment in the early stage, the middle stage and the late stage of the earthquake are obtained.

5. The method according to claim 1, characterized in that The obtaining of status data corresponding to different types of power equipment under earthquake conditions according to the group topology neural network specifically includes: Acquiring a layout structure of different types of power equipment in the converter station, and encoding the different types of power equipment into graph structure nodes in the group topology graph neural network according to the layout structure; Based on the group topology graph neural network, a graph attention network is used to model the propagation path of seismic waves to capture the vibration coupling effect between different types of power equipment; According to the vibration coupling effect, status data corresponding to different types of power equipment in an earthquake environment are obtained.

6. The method according to claim 1, characterized in that The step of inputting the state data and calculating the cluster centroid distance corresponding to each of the power devices according to the cluster statistical neural network specifically includes: Based on a clustering algorithm, the state data of each of the power devices is divided into a plurality of clusters, each of the clusters representing an abnormal state of the device; Calculating a centroid value corresponding to each of the clusters, where the centroid value is a value corresponding to an optimal center point of the state data in each of the clusters; The cluster centroid distance corresponding to each of the electrical devices is calculated based on the centroid value.

7. The method according to claim 6, characterized in that The calculating the cluster centroid distance corresponding to each of the power devices based on the centroid value specifically includes: The cluster centroid distance corresponding to each of the power devices is calculated based on the centroid value using the following formula: Among them, d i,c is the cluster centroid distance between the ith power device and the cth cluster, x i,j is the state data of the i-th power equipment in the j-th characteristic dimension, ω j is the weight coefficient corresponding to the jth feature dimension, n represents the number of feature dimensions, c j is the centroid value of the cth cluster on the jth feature dimension, σ j represents the standard deviation corresponding to the jth feature dimension, λ is the normalization coefficient, m represents the number of abnormal feature information in the parameter change trend information, represents the reference value corresponding to the kth abnormal feature information, and p is the sensitivity adjustment parameter.

8. The method according to claim 1, characterized in that The weighted calculation of the comprehensive status evaluation results corresponding to each of the power equipment in the converter station according to the cluster centroid distance specifically includes: Determine whether the cluster centroid distance is greater than a preset abnormal threshold; If the cluster centroid distance is greater than the preset abnormal threshold, confirming that the comprehensive status assessment result corresponding to the power equipment is an abnormal operating state; If the cluster centroid distance is less than or equal to the preset abnormal threshold, it is confirmed that the comprehensive status evaluation result corresponding to the power equipment is a normal operating state.

9. The method according to claim 8, characterized in that After confirming that the comprehensive status assessment result corresponding to the power equipment is an abnormal operating state if the cluster centroid distance is greater than the preset abnormal threshold, the method further includes: Obtaining an abnormality type corresponding to the power equipment according to the cluster centroid distance; Obtaining an equipment maintenance plan corresponding to the abnormality type from a preset abnormality handling plan database, wherein the preset abnormality handling plan database is used to store a correspondence between the abnormality type and the equipment maintenance plan; The equipment maintenance plan is sent to the user's terminal device.

10. A converter station earthquake status monitoring device based on clustering neural network, characterized in that: The device includes an acquisition module and a processing module, wherein: The acquisition module is used to respond to the user's operation of monitoring the operating status of different types of power equipment in the converter station under an earthquake environment, and obtain parameter information corresponding to each of the power equipment in the early stage, middle stage and late stage of the earthquake, respectively, wherein the parameter information includes vibration information, strain information, acceleration information and temperature information; perform time series analysis on the parameter information through a multi-scale damage characterization system, and obtain time series data corresponding to the power equipment in the early stage, middle stage and late stage of the earthquake; perform feature extraction operation on the time series data in a preset manner, and obtain parameter change trend information in the time series data based on the feature extraction operation, wherein the parameter change trend information includes vibration change trend information, strain change trend information, acceleration change trend information and temperature change trend information; according to the parameter change trend information, the parameter information corresponding to the early stage, middle stage and late stage of the earthquake, and according to the group topology graph neural network, obtain the status data corresponding to the different types of power equipment in the earthquake environment; The processing module is used to input the status data and calculate the cluster centroid distance corresponding to each of the power equipment according to the cluster statistical neural network; based on the cluster centroid distance, perform weighted calculation on the comprehensive status evaluation result corresponding to each of the power equipment in the converter station.

Images