Multi-dimensional space visualization transformer equipment monitoring and early warning system

Abstract

The invention discloses a multi-dimensional space visualization transformer equipment monitoring and early warning system, relates to the technical field of transformer equipment, and aims to solve the problem that more comprehensive monitoring and data early warning cannot be performed on a transformer. According to the invention, through coordinate axis confirmation and data type classification, the suitable chart type is selected to display the monitoring data, three-dimensional visualization of the data is realized, the visualization mode enables the data change trend and abnormal points to be clear at a glance, an operator is helped to rapidly understand and grasp the equipment state, and the operation efficiency is improved. And the target deployment transformer data after data fusion is imported into three-dimensional model software for scene creation, so that the direct display of the transformer equipment in a three-dimensional space is realized, the fastest transmission channel is automatically selected to ensure that the data can be quickly transmitted to a processing center from the equipment, the transmission delay is reduced, and the real-time response capability of the system is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of transformer equipment, and in particular to a transformer equipment monitoring and early warning system with multi-dimensional spatial visualization. Background Art

[0002] The transformer is designed for gas turbine generator sets. It is used to adjust the voltage generated by the generator to a voltage level suitable for grid transmission or user use.

[0003] The patent application with publication number CN117520976A discloses a transformer device using visual monitoring, which mainly divides the data field according to the abnormality of each cycle of the transformer's working condition information and the mutation degree between adjacent cycles, and obtains abnormal data points by performing outlier factor analysis based on the data field, thereby improving the accuracy and efficiency of abnormal data point detection. Although the above patent solves the problem of transformer monitoring, the following problems still exist in actual operation:

[0004] 1. The transformer and transformer monitoring equipment were not deployed in a more accurate position, resulting in the inability to accurately locate the abnormality later.

[0005] 2. The sensor data of the monitoring transformer is not monitored and analyzed in real time, resulting in the inability to issue timely alarms when the transformer is abnormal. Summary of the invention

[0006] The purpose of the present invention is to provide a transformer equipment monitoring and early warning system with multi-dimensional spatial visualization. By confirming the coordinate axis and classifying the data type, a suitable chart type is selected to display the monitoring data, thereby realizing three-dimensional visualization of the data. This visualization method makes the data change trend and abnormal points clear at a glance, which helps operators to quickly understand and grasp the equipment status. The target deployment transformer data completed by data fusion is imported into the three-dimensional model software for scene creation, thereby realizing the intuitive display of the transformer equipment in the three-dimensional space, automatically selecting the fastest transmission channel to ensure that the data can be quickly transmitted from the equipment to the processing center, reducing transmission delays, and improving the real-time response capability of the system, which can solve the problems in the prior art.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] Multi-dimensional spatial visualization of transformer equipment monitoring and early warning system, including:

[0009] Equipment location deployment unit, used to:

[0010] The transformer and the sensor are deployed at the location, and data is collected after the location deployment is completed. The data after the transformer and the sensor location deployment is completed are marked as the deployed transformer data;

[0011] Deploy data processing units for:

[0012] The deployment transformer data is transmitted to the processing center for data processing, and the target deployment transformer data is obtained after data processing;

[0013] Deploy data visualization building blocks for:

[0014] The target deployed transformer data is analyzed, and the analyzed data is fused. After the data fusion, a three-dimensional model is constructed, and after the three-dimensional model is constructed, a transformer visualization model is obtained.

[0015] Preferably, the device location deployment unit is further used for:

[0016] The transformer is a gas turbine power generation transformer, and the transformer equipment includes a main transformer, a distribution transformer and an auxiliary transformer, wherein the main transformer also includes a step-up transformer and a step-down transformer;

[0017] The transformer also includes monitoring sensors, including temperature sensors, oil level sensors, pressure sensors, current transformers, voltage transformers, insulation monitoring devices, gas analyzers, partial discharge detectors, and protection relays;

[0018] Deploy transformers and monitoring sensors at their respective locations;

[0019] Among them, the main transformer is installed near the generator; the distribution transformer is installed at the center of the electric load; the auxiliary transformer is installed near the electrical room;

[0020] The temperature sensor is installed on the oil tank or winding parts of the transformer; the oil level sensor is installed in the transformer oil tank; the pressure sensor is installed on the oil tank or pressure relief device; the current transformer is installed near the circuit breaker on the high-voltage side and the low-voltage side of the main transformer; the voltage transformer is installed on the high-voltage side; the insulation monitoring device is installed in the insulating oil system of the transformer; the gas analyzer is connected to the oil sampling system of the transformer; the partial discharge detector is installed near the bushing or winding of the transformer; the protection relay is installed in the control room or in the electrical room close to the transformer;

[0021] After the transformer and monitoring sensor locations are deployed, they are marked as deployed transformer data.

[0022] Preferably, the deployment data processing unit is further used for:

[0023] Transmitting deployment transformer data to a processing center;

[0024] When the deployment transformer data is transmitted, the transmission channel with the fastest transmission speed is automatically selected, wherein, according to the transmission amount of the deployment transformer data, the transmission channel whose channel remaining capacity is greater than the transmission amount of the deployment transformer data is selected as the transmission channel for transmission to the processing center;

[0025] After the transmission is completed, the deployed transformer data in the processing center will be processed;

[0026] Data processing includes dimensionality reduction, projection, graphical representation, and interactive data processing;

[0027] First, the deployed transformer data is subjected to dimensionality reduction processing. The dimensionality reduction processing is to select the data monitored by each sensor in the deployed transformer data. After the selection is completed, t-SNF is used to reduce the data dimension.

[0028] Map the deployed transformer data after dimensionality reduction to low-dimensional airborne, and project the time series data in the deployed transformer data onto the time axis;

[0029] The projected deployment transformer data is graphically represented. The graphical representation types include scatter plot, line graph, heat map and parallel coordinate graph. The scatter plot is a graph representing the relationship between two variables in the sensor data; the line graph is a graph showing the trend of time series data in the deployment transformer data; the heat map is a graph showing the spatial distribution of sensor data; and the parallel coordinate graph is a graph showing the coordinate axes of the transformer and sensor in the deployment transformer data.

[0030] The deployed transformer data is graphically represented and interactive charts are created using D3.js;

[0031] After the creation is completed, the target deployment transformer data is obtained.

[0032] Preferably, the deployment data visualization construction unit is further used for:

[0033] Perform data analysis on the target deployment transformer data, including global feature extraction and local feature extraction;

[0034] Global feature extraction is to perform data calculation on the target deployment transformer data, which includes average value calculation, standard layer calculation and median calculation. After the data calculation is completed, the overall feature data of the target deployment transformer data is obtained;

[0035] Analyze the time series feature data in the overall feature data. The time series features include periodicity, trend and seasonality.

[0036] Use Fourier transform technology to extract frequency feature data from time series feature data;

[0037] Local feature extraction is to calculate the local mean and variance of each data point in the target deployment transformer data, and obtain local feature data after calculation;

[0038] The local feature data is extracted using HHT technology;

[0039] Perform data fusion on frequency feature data and local feature data;

[0040] The data fusion process is to directly splice the frequency feature data and the local feature data in the feature dimension, and obtain the feature vector after the splicing is completed;

[0041] Use the attention mechanism to confirm the feature weights in the feature vector, and perform weighted fusion of the frequency feature data and the local feature data according to the feature weights;

[0042] The frequency feature data and local feature data completed by weighted fusion are used to train the machine learning model;

[0043] Finally, the target deployment transformer data is obtained after data fusion.

[0044] Preferably, the deployment data visualization construction unit is further used for:

[0045] Import the target deployment transformer data after data fusion into the 3D model software for scene creation;

[0046] After the scene is created, the transformer and sensor are used to create three-dimensional shapes;

[0047] Adjusting parameters of the created three-dimensional shape according to the frequency characteristic data and the local characteristic data, the parameters including size, position and rotation angle;

[0048] After the parameter adjustment of the three-dimensional shape is completed, the frequency characteristic data and the local characteristic data are mapped to the three-dimensional shape, and after the mapping is completed, the three-dimensional model of the transformer and the sensor is obtained;

[0049] The three-dimensional model is rendered using a rendering tool, and interactive functions are added to finally obtain a transformer visualization model.

[0050] Multi-dimensional spatial visualization of transformer equipment monitoring and early warning system, including:

[0051] Equipment data real-time monitoring unit, used for:

[0052] The sensor monitors the transformer in real time and collects the monitoring data. After the monitoring data collection is completed, the equipment monitoring data is obtained;

[0053] Monitoring data processing unit, used for:

[0054] The equipment monitoring data is transmitted to the processing center for data preprocessing, and the target monitoring data is obtained after the data preprocessing is completed;

[0055] Monitoring data visualization annotation and early warning unit, used for:

[0056] The target monitoring data is subjected to early warning analysis, and abnormal data is extracted based on the early warning analysis results. The extracted abnormal data and target monitoring data are subjected to three-dimensional visualization processing, and different degrees of warning alarms are issued according to the abnormal thresholds of the abnormal data.

[0057] Preferably, the device data real-time monitoring unit is also used for:

[0058] Sensors monitor transformers in real time;

[0059] Monitoring data include temperature data, oil level data, pressure data, current data, insulation data, gas analysis data, partial discharge data and protection relay status data;

[0060] After the monitoring data is acquired, it is uniquely coded and labeled;

[0061] The unique coding numbering process is to first confirm the coding rules, which are the coding length, the number of coding segments and the coding format;

[0062] After the coding rules are confirmed, the coding structure is designed;

[0063] The coding structure is designed to confirm the coding prefix of the monitoring data that has completed the coding rule confirmation. The coding prefix includes the sensor model, location code and data monitoring timestamp;

[0064] After the coding prefix is ​​confirmed, the equipment monitoring data is obtained.

[0065] Preferably, the monitoring data processing unit is further used for:

[0066] Transmit equipment monitoring data to the processing center for data preprocessing;

[0067] Furthermore, when the device monitors data for transmission, it automatically selects the transmission channel with the fastest transmission speed;

[0068] After the transmission is completed, the equipment monitoring data in the processing center will be pre-processed;

[0069] The data preprocessing process includes data cleaning, data conversion, feature extraction, data integration and data dimension reduction;

[0070] Data cleaning is to remove irrelevant and duplicate data from equipment monitoring data and process missing values ​​and outliers;

[0071] Data conversion is to perform data standardization and time series conversion on the equipment monitoring data after data cleaning;

[0072] Feature extraction is to extract feature data from the equipment monitoring data after data conversion;

[0073] Data integration is to merge and fuse the equipment monitoring data after feature extraction;

[0074] Data dimensionality reduction is to use PCA to reduce the dimension of the equipment monitoring data after data integration;

[0075] After data dimensionality reduction is completed, the target monitoring data is obtained.

[0076] Preferably, the monitoring data visualization annotation and early warning unit is also used for:

[0077] Compare the target monitoring data with the standard monitoring curve data, wherein the standard monitoring curve data is the qualified monitoring data threshold of the transformer when the sensor monitors the transformer, and the standard monitoring curve data is retrieved from the database;

[0078] Performing curve data conversion on each sub-data in the target monitoring data, wherein the sub-data is the monitoring data after data preprocessing;

[0079] Overlapping the sub-data after curve conversion with the standard curve data corresponding to the attributes in the standard monitoring curve data;

[0080] After the curves overlap, the sub-data that do not overlap with the standard monitoring curve data are marked as abnormal data;

[0081] The abnormal threshold of each abnormal data is confirmed, where the abnormal interval is the area of ​​the non-overlapping area.

[0082] Preferably, the monitoring data visualization annotation and early warning unit is also used for:

[0083] Confirm the coordinate axis of the target monitoring data, wherein the data type and chart type of the target monitoring data are first confirmed before confirming the coordinate axis;

[0084] Temperature data, oil level data, pressure data, current data, and insulation data are numerical data, so use line graphs or scatter graphs; protection relay status data are categorical data, so use bar graphs or pie charts; partial discharge data are event data, so use scatter graphs or time series graphs; gas analysis data are multivariate data, so use multi-axis charts;

[0085] After the data type and chart type of the target monitoring data are confirmed, the coordinate axes are confirmed. The coordinate axes are the X-axis and the Y-axis.

[0086] After the coordinate axis of the target monitoring data is confirmed, the three-dimensional visualization data of the target monitoring data is obtained;

[0087] Import the abnormal data in the target monitoring data into the axes corresponding to the 3D visualization data and distinguish them with special colors;

[0088] At the same time, the abnormal data is matched with the sensor. After the matching is completed, different degrees of sensor alarms are issued according to the abnormal threshold. The alarms are prompted in the display terminal and three-dimensional visualization data.

[0089] Compared with the prior art, the present invention has the following beneficial effects:

[0090] 1. The multi-dimensional space visualized transformer equipment monitoring and early warning system provided by the present invention can fully capture the overall and local characteristics of transformer data through global feature extraction and local feature extraction, ensure the integrity and accuracy of data analysis, and further extract the frequency characteristics and local characteristics of time series using Fourier transform and HHT technology. These advanced feature analyses help to discover hidden patterns and anomalies in the data, improve the prediction and diagnosis capabilities of the monitoring system, import the target deployment transformer data completed by data fusion into the three-dimensional model software for scene creation, and realize the intuitive display of transformer equipment in three-dimensional space. This visualization method helps operation and maintenance personnel to quickly understand the equipment layout and operation status.

[0091] 2. The transformer equipment monitoring and early warning system with multi-dimensional spatial visualization provided by the present invention realizes three-dimensional visualization of data by selecting appropriate chart types to display monitoring data through coordinate axis confirmation and data type classification. This visualization method makes data change trends and abnormal points clear at a glance, which helps operators to quickly understand and grasp the equipment status. Through multi-dimensional spatial visualization, the data changes of multiple monitoring parameters, as well as the correlation and mutual influence between them, can be displayed simultaneously. This multi-dimensional early warning analysis helps to understand the operating status and potential risks of the equipment more comprehensively and deeply, and improves the accuracy and effectiveness of the early warning system.

[0092] 3. The transformer equipment monitoring and early warning system with multi-dimensional spatial visualization provided by the present invention automatically selects the fastest transmission channel to ensure that data can be quickly transmitted from the equipment to the processing center, reduces transmission delays, and improves the real-time response capability of the system. The data cleaning step effectively removes irrelevant data and duplicate data, and processes missing values ​​and outliers, significantly improving the quality and reliability of the data and laying a solid foundation for subsequent analysis. The data processing process includes dimensionality reduction, projection, graphical representation, and interactive processing. The orderly combination of these steps enables a large amount of complex transformer data to be effectively simplified and visualized. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] Figure 1 A schematic diagram of a transformer equipment monitoring and early warning multi-dimensional space visualization unit of the present invention;

[0094] Figure 2 This is a schematic diagram of the multi-dimensional spatial visualization process of transformer equipment monitoring and early warning of the present invention. DETAILED DESCRIPTION

[0095] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0096] In order to solve the problem that the transformer and transformer monitoring equipment are not deployed in a more accurate position in the existing technology, which leads to the inability to accurately locate the abnormality later, please refer to Figure 1 and Figure 2 , this embodiment provides the following technical solutions:

[0097] Multi-dimensional spatial visualization of transformer equipment monitoring and early warning system, including:

[0098] Equipment location deployment unit, used to:

[0099] The transformer and the sensor are deployed at the location, and data is collected after the location deployment is completed. The data after the transformer and the sensor location deployment is completed are marked as the deployed transformer data;

[0100] Deploy data processing units for:

[0101] The deployment transformer data is transmitted to the processing center for data processing, and the target deployment transformer data is obtained after data processing;

[0102] Deploy data visualization building blocks for:

[0103] The target deployed transformer data is analyzed, and the analyzed data is fused. After the data fusion, a three-dimensional model is constructed, and after the three-dimensional model is constructed, a transformer visualization model is obtained.

[0104] Specifically, by deploying units at equipment locations, the maintenance costs and downtime losses caused by equipment failures can be reduced. By deploying data processing units, the data dimension can be greatly reduced while retaining key information, the calculation complexity can be reduced, and the processing efficiency can be improved. By deploying data visualization construction units, the operating status and key indicators of the transformer can be displayed in real time, helping operation and maintenance personnel to quickly discover potential problems and take corresponding measures.

[0105] The device location deployment unit is also used to:

[0106] The transformer is a gas turbine power generation transformer, and the transformer equipment includes a main transformer, a distribution transformer and an auxiliary transformer, wherein the main transformer also includes a step-up transformer and a step-down transformer;

[0107] The transformer also includes monitoring sensors, including temperature sensors, oil level sensors, pressure sensors, current transformers, voltage transformers, insulation monitoring devices, gas analyzers, partial discharge detectors, and protection relays;

[0108] Deploy transformers and monitoring sensors at their respective locations;

[0109] Among them, the main transformer is installed near the generator; the distribution transformer is installed at the center of the electric load; the auxiliary transformer is installed near the electrical room;

[0110] The temperature sensor is installed on the oil tank or winding parts of the transformer; the oil level sensor is installed in the transformer oil tank; the pressure sensor is installed on the oil tank or pressure relief device; the current transformer is installed near the circuit breaker on the high-voltage side and the low-voltage side of the main transformer; the voltage transformer is installed on the high-voltage side; the insulation monitoring device is installed in the insulating oil system of the transformer; the gas analyzer is connected to the oil sampling system of the transformer; the partial discharge detector is installed near the bushing or winding of the transformer; the protection relay is installed in the control room or in the electrical room close to the transformer;

[0111] After the transformer and monitoring sensor locations are deployed, they are marked as deployed transformer data.

[0112] Specifically, by installing monitoring sensors (such as temperature sensors, oil level sensors, etc.) directly on key parts of the transformer, the operating status of the transformer can be monitored in real time. Once an abnormal situation occurs, the system can quickly issue an alarm, greatly shortening the fault response time and reducing the risk of accidents. Advanced sensors such as insulation monitoring devices, gas analyzers and partial discharge detectors can capture subtle changes inside the transformer. Through data analysis, potential faults can be predicted in advance, preventive maintenance can be achieved, and losses caused by sudden failures can be avoided. Through multi-dimensional spatial visualization technology, operators can intuitively see the distribution and real-time data of the transformer and its monitoring sensors, improve operational efficiency and accuracy, and reduce the risk of misoperation. Through accurate prediction and preventive maintenance, maintenance costs and downtime losses caused by equipment failures can be reduced. At the same time, optimizing maintenance plans and improving operational efficiency also help reduce overall operation and maintenance costs.

[0113] Deployment of data processing units, also used for:

[0114] Transmitting deployment transformer data to a processing center;

[0115] When the deployment transformer data is transmitted, the transmission channel with the fastest transmission speed is automatically selected, wherein, according to the transmission amount of the deployment transformer data, the transmission channel whose channel remaining capacity is greater than the transmission amount of the deployment transformer data is selected as the transmission channel for transmission to the processing center;

[0116] After the transmission is completed, the deployed transformer data in the processing center will be processed;

[0117] Data processing includes dimensionality reduction, projection, graphical representation, and interactive data processing;

[0118] First, the deployed transformer data is subjected to dimensionality reduction processing. The dimensionality reduction processing is to select the data monitored by each sensor in the deployed transformer data. After the selection is completed, t-SNF is used to reduce the data dimension.

[0119] Map the deployed transformer data after dimensionality reduction to low-dimensional airborne, and project the time series data in the deployed transformer data onto the time axis;

[0120] The projected deployment transformer data is graphically represented. The graphical representation types include scatter plot, line graph, heat map and parallel coordinate graph. The scatter plot is a graph representing the relationship between two variables in the sensor data; the line graph is a graph showing the trend of time series data in the deployment transformer data; the heat map is a graph showing the spatial distribution of sensor data; and the parallel coordinate graph is a graph showing the coordinate axes of the transformer and sensor in the deployment transformer data.

[0121] The deployed transformer data is graphically represented and interactive charts are created using D3.js;

[0122] After the creation is completed, the target deployment transformer data is obtained.

[0123] Specifically, the channel with the fastest transmission speed is automatically selected, and the remaining capacity of the channel is matched according to the transmission volume of the deployed transformer data to ensure the efficiency and stability of data transmission. This adaptive channel selection mechanism effectively reduces the delay and congestion of data transmission, and improves the response speed and reliability of the entire system. The data processing process includes dimensionality reduction, projection, graphic representation and interactive processing. The orderly combination of these steps enables a large amount of complex transformer data to be effectively simplified and visualized, which is convenient for subsequent analysis and monitoring. Using advanced algorithms such as t-SNF for data dimensionality reduction can significantly reduce data dimensions while retaining key information, reduce computational complexity, and improve processing efficiency. This is particularly important for processing massive amounts of transformer monitoring data. It provides a variety of graphical representations (scatter plots, line graphs, heat maps, parallel coordinate graphs), each of which is optimized for different features of transformer data, allowing monitoring personnel to intuitively understand the transformer's operating status, fault trends, spatial distribution, and the relationship between multiple variables, thereby improving the accuracy and comprehensiveness of monitoring. Through real-time monitoring and visual analysis of transformer data, the system can promptly detect abnormal data and potential faults, provide early warning information to operation and maintenance personnel, improve fault response speed and diagnostic accuracy, and reduce power outages and economic losses caused by transformer failures.

[0124] Deployment of data visualization building blocks, also used for:

[0125] Perform data analysis on the target deployment transformer data, including global feature extraction and local feature extraction;

[0126] Global feature extraction is to perform data calculation on the target deployment transformer data, which includes average value calculation, standard layer calculation and median calculation. After the data calculation is completed, the overall feature data of the target deployment transformer data is obtained;

[0127] Analyze the time series feature data in the overall feature data. The time series features include periodicity, trend and seasonality.

[0128] Use Fourier transform technology to extract frequency feature data from time series feature data;

[0129] Local feature extraction is to calculate the local mean and variance of each data point in the target deployment transformer data, and obtain local feature data after calculation;

[0130] The local feature data is extracted using HHT technology;

[0131] Perform data fusion on frequency feature data and local feature data;

[0132] The data fusion process is to directly splice the frequency feature data and the local feature data in the feature dimension, and obtain the feature vector after the splicing is completed;

[0133] Use the attention mechanism to confirm the feature weights in the feature vector, and perform weighted fusion of the frequency feature data and the local feature data according to the feature weights;

[0134] The frequency feature data and local feature data completed by weighted fusion are used to train the machine learning model;

[0135] Finally, the target deployment transformer data is obtained after data fusion.

[0136] Import the target deployment transformer data after data fusion into the 3D model software for scene creation;

[0137] After the scene is created, the transformer and sensor are used to create three-dimensional shapes;

[0138] Adjusting parameters of the created three-dimensional shape according to the frequency characteristic data and the local characteristic data, the parameters including size, position and rotation angle;

[0139] After the parameter adjustment of the three-dimensional shape is completed, the frequency characteristic data and the local characteristic data are mapped to the three-dimensional shape, and after the mapping is completed, the three-dimensional model of the transformer and the sensor is obtained;

[0140] The three-dimensional model is rendered using a rendering tool, and interactive functions are added to finally obtain a transformer visualization model.

[0141] Specifically, through global feature extraction (such as mean, standard deviation, median, etc.) and local feature extraction (such as local mean, variance, etc.), the overall and local characteristics of transformer data can be fully captured to ensure the integrity and accuracy of data analysis. Fourier transform and HHT technology are used to further extract frequency characteristics and local characteristics of time series. These advanced feature analyses help to discover hidden patterns and anomalies in the data and improve the prediction and diagnosis capabilities of the monitoring system. The frequency feature data and local feature data are spliced ​​in the feature dimension and weighted fused through the attention mechanism. This method can automatically learn and optimize the importance of different features, improve the efficiency and accuracy of data fusion, and the fused feature vector contains rich transformer status information, which provides strong data support for subsequent machine learning model training. The target deployment transformer data after data fusion is imported into the three-dimensional model software for scene creation, realizing the intuitive display of transformer equipment in three-dimensional space. This visualization method helps operation and maintenance personnel to quickly understand the equipment layout and operating status. Adjusting the parameters of the three-dimensional shape (such as size, position, and rotation angle) according to the frequency characteristic data and local characteristic data can more accurately reflect the actual status of the transformer and improve the accuracy and reliability of the monitoring system. By adding rendering tools and interactive functions, the transformer visualization model not only has realistic visual effects, but also allows users to perform interactive operations such as rotation, zooming, and viewing detailed information. This interactivity enhances the user experience and improves the ease of use and practicality of the monitoring system. It can display the operating status and key indicators of the transformer in real time, helping operation and maintenance personnel to quickly discover potential problems and take corresponding measures. At the same time, through data analysis and model prediction, it can also provide strong support for decision-making, improve operation and maintenance efficiency and the scientific nature of decision-making.

[0142] In order to solve the problem in the prior art that the sensor data of the monitoring transformer is not monitored and analyzed in real time, resulting in the failure to issue an alarm in time when the transformer is abnormal, please refer to Figure 1 and Figure 2 , this embodiment provides the following technical solutions:

[0143] Multi-dimensional spatial visualization of transformer equipment monitoring and early warning system, including:

[0144] Equipment data real-time monitoring unit, used for:

[0145] The sensor monitors the transformer in real time and collects the monitoring data. After the monitoring data collection is completed, the equipment monitoring data is obtained;

[0146] Monitoring data processing unit, used for:

[0147] The equipment monitoring data is transmitted to the processing center for data preprocessing, and the target monitoring data is obtained after the data preprocessing is completed;

[0148] Monitoring data visualization annotation and early warning unit, used for:

[0149] The target monitoring data is subjected to early warning analysis, and abnormal data is extracted based on the early warning analysis results. The extracted abnormal data and target monitoring data are subjected to three-dimensional visualization processing, and different degrees of warning alarms are issued according to the abnormal thresholds of the abnormal data.

[0150] Specifically, the real-time monitoring unit of equipment data can ensure that each piece of data has a unique identifier, avoiding data confusion or duplication. The monitoring data processing unit takes into account the balance between efficiency and stability, ensuring the stable operation of the system under high load. The monitoring data visualization annotation and early warning unit can adjust the monitoring parameters, chart types, and alarm rules according to actual needs.

[0151] The equipment data real-time monitoring unit is also used for:

[0152] Sensors monitor transformers in real time;

[0153] Monitoring data include temperature data, oil level data, pressure data, current data, insulation data, gas analysis data, partial discharge data and protection relay status data;

[0154] After the monitoring data is acquired, it is uniquely coded and labeled;

[0155] The unique coding numbering process is to first confirm the coding rules, which are the coding length, the number of coding segments and the coding format;

[0156] After the coding rules are confirmed, the coding structure is designed;

[0157] The coding structure is designed to confirm the coding prefix of the monitoring data that has completed the coding rule confirmation. The coding prefix includes the sensor model, location code and data monitoring timestamp;

[0158] After the coding prefix is ​​confirmed, the equipment monitoring data is obtained.

[0159] Specifically, by uniquely coding and labeling the monitoring data (such as temperature, oil level, pressure, current, insulation status, gas analysis, partial discharge, protection relay status, etc.), it can be ensured that each piece of data has a unique identifier, avoiding data confusion or duplication, and improving the accuracy of data management. The coding prefix contains the sensor model, location code and data monitoring timestamp, which provides strong support for data traceability. When a problem occurs in the transformer, the specific sensor, location and time period can be quickly located to facilitate the analysis of the cause of the fault. In the multi-dimensional space visualization system, these data can be intuitively displayed in the form of charts, images, etc., so that the operating status of the transformer is clear at a glance. Once the data is abnormal, the system can immediately issue an early warning to prompt the maintenance personnel to pay attention.

[0160] The monitoring data processing unit is also used for:

[0161] Transmit equipment monitoring data to the processing center for data preprocessing;

[0162] Furthermore, when the device monitors data for transmission, it automatically selects the transmission channel with the fastest transmission speed;

[0163] After the transmission is completed, the equipment monitoring data in the processing center will be pre-processed;

[0164] The data preprocessing process includes data cleaning, data conversion, feature extraction, data integration and data dimension reduction;

[0165] Data cleaning is to remove irrelevant and duplicate data from equipment monitoring data and process missing values ​​and outliers;

[0166] Data conversion is to perform data standardization and time series conversion on the equipment monitoring data after data cleaning;

[0167] Feature extraction is to extract feature data from the equipment monitoring data after data conversion;

[0168] Data integration is to merge and fuse the equipment monitoring data after feature extraction;

[0169] Data dimensionality reduction is to use PCA to reduce the dimension of the equipment monitoring data after data integration;

[0170] After data dimensionality reduction is completed, the target monitoring data is obtained.

[0171] Specifically, the fastest transmission channel is automatically selected to ensure that data can be quickly transmitted from the device to the processing center, reducing transmission delays and improving the real-time response capability of the system. The data cleaning step effectively removes irrelevant data and duplicate data, and handles missing values ​​and outliers, significantly improving the quality and reliability of the data, laying a solid foundation for subsequent analysis. Data standardization enables data from different sources and dimensions to be compared and analyzed on the same platform, increasing the comparability and availability of data. Feature extraction can mine features that are critical to analysis from the converted data, which helps to more accurately identify the status of the equipment and potential problems. The data integration step enriches the data dimension of the analysis by merging and integrating information from multiple data sources, enabling the early warning system to evaluate the status of the equipment more comprehensively and comprehensively. PCA is used for data dimensionality reduction, which not only reduces the complexity of data processing, but also retains the main information of the data, facilitating subsequent data analysis and visualization. When the reduced-dimensional data is visualized in multidimensional space, it can more clearly show the changing trend and abnormal pattern of the equipment status, improving the intuitiveness and accuracy of the early warning. The design of the entire data transmission and preprocessing process takes into account the balance between efficiency and stability, ensuring the stable operation of the system under high load.

[0172] The monitoring data visualization annotation and early warning unit is also used for:

[0173] Compare the target monitoring data with the standard monitoring curve data, wherein the standard monitoring curve data is the qualified monitoring data threshold of the transformer when the sensor monitors the transformer, and the standard monitoring curve data is retrieved from the database;

[0174] Performing curve data conversion on each sub-data in the target monitoring data, wherein the sub-data is the monitoring data after data preprocessing;

[0175] Overlapping the sub-data after curve conversion with the standard curve data corresponding to the attributes in the standard monitoring curve data;

[0176] After the curves overlap, the sub-data that do not overlap with the standard monitoring curve data are marked as abnormal data;

[0177] The abnormal threshold of each abnormal data is confirmed, where the abnormal interval is the area of ​​the non-overlapping area.

[0178] Confirm the coordinate axis of the target monitoring data, wherein the data type and chart type of the target monitoring data are first confirmed before confirming the coordinate axis;

[0179] Temperature data, oil level data, pressure data, current data, and insulation data are numerical data, so use line graphs or scatter graphs; protection relay status data are categorical data, so use bar graphs or pie charts; partial discharge data are event data, so use scatter graphs or time series graphs; gas analysis data are multivariate data, so use multi-axis charts;

[0180] After the data type and chart type of the target monitoring data are confirmed, the coordinate axes are confirmed. The coordinate axes are the X-axis and the Y-axis.

[0181] After the coordinate axis of the target monitoring data is confirmed, the three-dimensional visualization data of the target monitoring data is obtained;

[0182] Import the abnormal data in the target monitoring data into the axes corresponding to the 3D visualization data and distinguish them with special colors;

[0183] At the same time, the abnormal data is matched with the sensor. After the matching is completed, different degrees of sensor alarms are issued according to the abnormal threshold. The alarms are prompted in the display terminal and three-dimensional visualization data.

[0184] Specifically, by comparing the target monitoring data with the standard monitoring curve data, the monitoring data that exceeds the normal range or is abnormal can be quickly identified. This method is more accurate and efficient than traditional manual comparison or single threshold judgment, reduces human errors, and improves the accuracy of detection. By confirming the coordinate axis and classifying the data type, the appropriate chart type (such as line chart, scatter chart, bar chart, pie chart, time series chart, multi-axis chart, etc.) is selected to display the monitoring data, realizing the three-dimensional visualization of the data. This visualization method makes the data change trend and abnormal points clear at a glance, which helps operators to quickly understand and grasp the equipment status, mark the sub-data that does not overlap with the standard monitoring curve as abnormal data, and calculate the abnormal threshold (that is, the area of ​​the non-overlapping area), so as to realize the accurate positioning and quantification of the abnormal data. This helps to more accurately assess the degree and risk of equipment failure, provide a basis for subsequent processing and decision-making, correspond abnormal data with sensors, and issue different degrees of sensor alarms according to the abnormal threshold. This real-time alarm mechanism can quickly notify relevant personnel to pay attention to and deal with abnormal situations, avoid further expansion and deterioration of equipment failures, and improve the safety and reliability of equipment. Through multi-dimensional spatial visualization, it can simultaneously display the data changes of multiple monitoring parameters, as well as their associations and mutual influences. This multi-dimensional early warning analysis helps to understand the operating status and potential risks of equipment more comprehensively and deeply, and improves the accuracy and effectiveness of the early warning system. By marking abnormal data in special colors in the three-dimensional visualization data and issuing alarm prompts in the display terminal and three-dimensional visualization data, users can intuitively see abnormal situations and equipment status. This design enhances the user experience and improves user satisfaction and trust in the early warning system.

[0185] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0186] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

Claims

1. A multi-dimensional spatial visualization transformer equipment monitoring and early warning system, characterized in that: include: Equipment location deployment unit, used to: The transformer and the sensor are deployed at the location, and data is collected after the location deployment is completed. The data after the transformer and the sensor location deployment is completed are marked as the deployed transformer data; Deploy data processing units for: The deployment transformer data is transmitted to the processing center for data processing, and the target deployment transformer data is obtained after data processing; Deploy data visualization building blocks for: The target deployed transformer data is analyzed, and the analyzed data is fused. After the data fusion, a three-dimensional model is constructed, and after the three-dimensional model is constructed, a transformer visualization model is obtained.

2. The transformer equipment monitoring and early warning system with multi-dimensional spatial visualization according to claim 1 is characterized in that: The device location deployment unit is further used for: The transformer is a gas turbine power generation transformer, and the transformer equipment includes a main transformer, a distribution transformer and an auxiliary transformer, wherein the main transformer also includes a step-up transformer and a step-down transformer; The transformer also includes monitoring sensors, including temperature sensors, oil level sensors, pressure sensors, current transformers, voltage transformers, insulation monitoring devices, gas analyzers, partial discharge detectors, and protection relays; Deploy transformers and monitoring sensors at their respective locations; Among them, the main transformer is installed near the generator; the distribution transformer is installed at the center of the electric load; the auxiliary transformer is installed near the electrical room; The temperature sensor is installed on the oil tank or winding parts of the transformer; the oil level sensor is installed in the transformer oil tank; the pressure sensor is installed on the oil tank or pressure relief device; the current transformer is installed near the circuit breaker on the high-voltage side and the low-voltage side of the main transformer; the voltage transformer is installed on the high-voltage side; the insulation monitoring device is installed in the insulating oil system of the transformer; the gas analyzer is connected to the oil sampling system of the transformer; the partial discharge detector is installed near the bushing or winding of the transformer; the protection relay is installed in the control room or in the electrical room close to the transformer; After the transformer and monitoring sensor locations are deployed, they are marked as deployed transformer data.

3. The transformer equipment monitoring and early warning system with multi-dimensional spatial visualization according to claim 2 is characterized in that: The deployment data processing unit is further used for: Transmitting deployment transformer data to a processing center; When the deployment transformer data is transmitted, the transmission channel with the fastest transmission speed is automatically selected, wherein, according to the transmission amount of the deployment transformer data, the transmission channel whose channel remaining capacity is greater than the transmission amount of the deployment transformer data is selected as the transmission channel for transmission to the processing center; After the transmission is completed, the deployed transformer data in the processing center will be processed; Data processing includes dimensionality reduction, projection, graphical representation, and interactive data processing; First, the deployed transformer data is subjected to dimensionality reduction processing. The dimensionality reduction processing is to select the data monitored by each sensor in the deployed transformer data. After the selection is completed, t-SNF is used to reduce the data dimension. Map the deployed transformer data after dimensionality reduction to low-dimensional airborne, and project the time series data in the deployed transformer data onto the time axis; The projected deployment transformer data is graphically represented. The graphical representation types include scatter plot, line graph, heat map and parallel coordinate graph. The scatter plot is a graph representing the relationship between two variables in the sensor data; the line graph is a graph showing the trend of time series data in the deployment transformer data; the heat map is a graph showing the spatial distribution of sensor data; and the parallel coordinate graph is a graph showing the coordinate axes of the transformer and sensor in the deployment transformer data. The deployed transformer data is graphically represented and interactive charts are created using D3.js; After the creation is completed, the target deployment transformer data is obtained.

4. The transformer equipment monitoring and early warning system with multi-dimensional spatial visualization according to claim 3 is characterized in that: The deployment data visualization construction unit is also used for: Perform data analysis on the target deployment transformer data, including global feature extraction and local feature extraction; Global feature extraction is to perform data calculation on the target deployment transformer data, which includes average value calculation, standard layer calculation and median calculation. After the data calculation is completed, the overall feature data of the target deployment transformer data is obtained; Analyze the time series feature data in the overall feature data. The time series features include periodicity, trend and seasonality. Use Fourier transform technology to extract frequency feature data from time series feature data; Local feature extraction is to calculate the local mean and variance of each data point in the target deployment transformer data, and obtain local feature data after calculation; The local feature data is extracted using HHT technology; Perform data fusion on frequency feature data and local feature data; The data fusion process is to directly splice the frequency feature data and the local feature data in the feature dimension, and obtain the feature vector after the splicing is completed; Use the attention mechanism to confirm the feature weights in the feature vector, and perform weighted fusion of the frequency feature data and the local feature data according to the feature weights; The frequency feature data and local feature data completed by weighted fusion are used to train the machine learning model; Finally, the target deployment transformer data is obtained after data fusion.

5. The transformer equipment monitoring and early warning system with multi-dimensional spatial visualization according to claim 4 is characterized in that: The deployment data visualization construction unit is also used for: Import the target deployment transformer data after data fusion into the 3D model software for scene creation; After the scene is created, the transformer and sensor are used to create three-dimensional shapes; Adjusting parameters of the created three-dimensional shape according to the frequency characteristic data and the local characteristic data, the parameters including size, position and rotation angle; After the parameter adjustment of the three-dimensional shape is completed, the frequency characteristic data and the local characteristic data are mapped to the three-dimensional shape, and after the mapping is completed, the three-dimensional model of the transformer and the sensor is obtained; The three-dimensional model is rendered using a rendering tool, and interactive functions are added to finally obtain a transformer visualization model.

6. The transformer equipment monitoring and early warning system with multi-dimensional spatial visualization according to claim 4 is characterized in that: Also includes: Equipment data real-time monitoring unit, used for: The sensor monitors the transformer in real time and collects the monitoring data. After the monitoring data collection is completed, the equipment monitoring data is obtained; Monitoring data processing unit, used for: The equipment monitoring data is transmitted to the processing center for data preprocessing, and the target monitoring data is obtained after the data preprocessing is completed; Monitoring data visualization annotation and early warning unit, used for: The target monitoring data is subjected to early warning analysis, and abnormal data is extracted based on the early warning analysis results. The extracted abnormal data and target monitoring data are subjected to three-dimensional visualization processing, and different degrees of warning alarms are issued according to the abnormal thresholds of the abnormal data.

7. The transformer equipment monitoring and early warning system with multi-dimensional spatial visualization according to claim 6 is characterized in that: The device data real-time monitoring unit is also used for: Sensors monitor transformers in real time; Monitoring data include temperature data, oil level data, pressure data, current data, insulation data, gas analysis data, partial discharge data and protection relay status data; After the monitoring data is acquired, it is uniquely coded and labeled; The unique coding numbering process is to first confirm the coding rules, which are the coding length, the number of coding segments and the coding format; After the coding rules are confirmed, the coding structure is designed; The coding structure is designed to confirm the coding prefix of the monitoring data that has completed the coding rule confirmation. The coding prefix includes the sensor model, location code and data monitoring timestamp; After the coding prefix is ​​confirmed, the equipment monitoring data is obtained.

8. The transformer equipment monitoring and early warning system with multi-dimensional spatial visualization according to claim 7 is characterized in that: The monitoring data processing unit is further used for: Transmit equipment monitoring data to the processing center for data preprocessing; Furthermore, when the device monitors data for transmission, it automatically selects the transmission channel with the fastest transmission speed; After the transmission is completed, the equipment monitoring data in the processing center will be pre-processed; The data preprocessing process includes data cleaning, data conversion, feature extraction, data integration and data dimension reduction; Data cleaning is to remove irrelevant and duplicate data from equipment monitoring data and process missing values ​​and outliers; Data conversion is to perform data standardization and time series conversion on the equipment monitoring data after data cleaning; Feature extraction is to extract feature data from the equipment monitoring data after data conversion; Data integration is to merge and fuse the equipment monitoring data after feature extraction; Data dimensionality reduction is to use PCA to reduce the dimension of the equipment monitoring data after data integration; After data dimensionality reduction is completed, the target monitoring data is obtained.

9. The transformer equipment monitoring and early warning system with multi-dimensional spatial visualization according to claim 8 is characterized in that: The monitoring data visualization labeling and early warning unit is also used for: Compare the target monitoring data with the standard monitoring curve data, wherein the standard monitoring curve data is the qualified monitoring data threshold of the transformer when the sensor monitors the transformer, and the standard monitoring curve data is retrieved from the database; Performing curve data conversion on each sub-data in the target monitoring data, wherein the sub-data is the monitoring data after data preprocessing; Overlapping the sub-data after curve conversion with the standard curve data corresponding to the attributes in the standard monitoring curve data; After the curves overlap, the sub-data that do not overlap with the standard monitoring curve data are marked as abnormal data; The abnormal threshold of each abnormal data is confirmed, where the abnormal interval is the area of ​​the non-overlapping area.

10. The transformer equipment monitoring and early warning system with multi-dimensional spatial visualization according to claim 9 is characterized in that: The monitoring data visualization annotation and early warning unit is also used for: Confirm the coordinate axis of the target monitoring data, wherein the data type and chart type of the target monitoring data are first confirmed before confirming the coordinate axis; Temperature data, oil level data, pressure data, current data, and insulation data are numerical data, so use line graphs or scatter graphs; protection relay status data are categorical data, so use bar graphs or pie charts; partial discharge data are event data, so use scatter graphs or time series graphs; gas analysis data are multivariate data, so use multi-axis charts; After the data type and chart type of the target monitoring data are confirmed, the coordinate axes are confirmed. The coordinate axes are the X-axis and the Y-axis. After the coordinate axis of the target monitoring data is confirmed, the three-dimensional visualization data of the target monitoring data is obtained; Import the abnormal data in the target monitoring data into the axes corresponding to the 3D visualization data and distinguish them with special colors; At the same time, the abnormal data is matched with the sensor. After the matching is completed, different degrees of sensor alarms are issued according to the abnormal threshold. The alarms are prompted in the display terminal and three-dimensional visualization data.

Images