A method and system for visualizing operation and maintenance data based on smart grids

Abstract

This invention relates to the field of data visualization technology, and provides a method and system for visualizing operation and maintenance data based on smart grids. This method aims to improve the efficiency and intelligence of power grid operation and maintenance visualization, thereby enhancing the timeliness of power grid operation and maintenance decision-making and management. The method includes: real-time acquisition of multi-source heterogeneous operation and maintenance data sets from the smart grid; cross-modal cleaning and time alignment processing of the operation and maintenance data sets to generate a preprocessed data set; dynamic feature fusion and correlation analysis of the preprocessed data set to generate a set of equipment status features and a set of equipment-environment interaction features; generation of an interactive visualization element set based on the set of equipment status features and the set of equipment-environment interaction features through a multi-dimensional visualization strategy; dynamic rendering and layout optimization of the interactive visualization element set to output a real-time updated power grid operation and maintenance visualization interface; and configuration of a user operation priority response mechanism for handling interaction conflicts within the power grid operation and maintenance visualization interface.

Description

Technical Field

[0001] This invention belongs to the field of data visualization technology, specifically relating to a method and system for visualizing operation and maintenance data based on smart grids. Background Technology

[0002] With the rapid development of smart grids, grid data is exhibiting characteristics of multi-source heterogeneity. Grid data visualization technology has emerged to address this, aiming to present complex data in an intuitive graphical way, helping operation and maintenance personnel better understand the grid's operational status. Visualization methods can improve operation and maintenance efficiency and promptly identify potential problems.

[0003] However, existing power grid data visualization technologies have many shortcomings: in terms of data processing, there is a lack of effective cleaning and time alignment mechanisms for multi-source heterogeneous operation and maintenance data, resulting in inconsistent data quality and affecting the accuracy of subsequent analysis; in terms of feature extraction, the deep relationship between equipment status and environmental interaction is not fully explored, making it difficult to comprehensively reflect the actual situation of the power grid; the existing visualization process has a single interaction method, and the visualization elements lack dynamism and real-time update capabilities, which cannot meet the complex and ever-changing operational needs of operation and maintenance personnel.

[0004] In view of this, how to improve the efficiency and intelligence of power grid operation and maintenance visualization in order to enhance the timeliness of power grid operation and maintenance decision-making and management is a technical problem that needs to be solved at present. Summary of the Invention

[0005] This invention provides a method and system for visualizing operation and maintenance data based on smart grids, which improves the efficiency and intelligence of grid operation and maintenance visualization, thereby enhancing the timeliness of grid operation and maintenance decision-making and management.

[0006] In a first aspect, embodiments of the present invention provide a method for visualizing operation and maintenance data based on a smart grid, applied to an operation and maintenance data visualization system. The method includes: real-time acquisition of a multi-source heterogeneous set of operation and maintenance data in the smart grid; performing cross-modal cleaning and time alignment processing on the operation and maintenance data set to generate a preprocessed data set; performing dynamic feature fusion and correlation analysis on the preprocessed data set to generate a set of equipment status features and a set of equipment-environment interaction features; generating a set of interactive visualization elements based on the set of equipment status features and the set of equipment-environment interaction features through a multi-dimensional visualization strategy; dynamically rendering and optimizing the layout of the set of interactive visualization elements to output a real-time updated power grid operation and maintenance visualization interface; wherein the power grid operation and maintenance visualization interface is configured with a user operation priority response mechanism for handling interaction conflicts.

[0007] Secondly, embodiments of the present invention provide an operation and maintenance data visualization system, which includes a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the above-described method.

[0008] Thirdly, embodiments of the present invention provide a computer-readable storage medium including a computer program, which, when run on an operation and maintenance data visualization system, causes the operation and maintenance data visualization system to perform the steps of the above-described method.

[0009] In this invention, by collecting multi-source heterogeneous operation and maintenance data sets from the smart grid in real time, various relevant information can be comprehensively acquired. Cross-modal cleaning and time alignment processing effectively improve data quality and consistency, making the preprocessed data set more usable. Dynamic feature fusion and correlation analysis can deeply mine potential relationships in the data, thereby generating a feature set that accurately reflects equipment status and equipment-environment interactions. Furthermore, based on a multi-dimensional visualization strategy, an interactive visualization element set is generated, providing users with diverse interaction methods. Finally, dynamic rendering and layout optimization ensure real-time updates and a reasonable layout of the visualization interface. In addition, configuring a user operation priority response mechanism ensures orderly responses to user operations in complex interaction scenarios. Thus, the efficiency and intelligence of power grid operation and maintenance visualization can be improved, thereby enhancing the timeliness of power grid operation and maintenance decision-making and management. Attached Figure Description

[0010] Figure 1 This is a flowchart illustrating a smart grid-based operation and maintenance data visualization method provided in an embodiment of the present invention.

[0011] Figure 2 This is a schematic diagram of the structure of an operation and maintenance data visualization system provided in an embodiment of the present invention. Detailed Implementation

[0012] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the technical solutions of the present invention, and not all embodiments. Based on the embodiments described in this document, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the technical solutions of the present invention.

[0013] See Figure 1 This is a method for visualizing operation and maintenance data based on a smart grid, provided in an embodiment of the present invention. This method can be applied to an operation and maintenance data visualization system, and the specific process is as shown in S110-S150.

[0014] S110: Real-time collection of multi-source heterogeneous operation and maintenance data sets in smart grids.

[0015] In this embodiment of the invention, the smart grid comprises numerous different types of power equipment. To comprehensively understand the grid's operational status, it is necessary to collect multi-source, heterogeneous operation and maintenance data sets in real time. For example, data can be collected from multiple power devices distributed across various regions.

[0016] Regarding current time series data, taking a large substation as an example, it has multiple transmission lines, each with corresponding current data. This current data exists in the form of a time series. For example, within one minute, the recorded current values ​​are 50A, 52A, 51A, etc., forming the initial current time series data.

[0017] Optionally, the voltage data segment is obtained from voltage monitoring devices connected to different nodes. These devices record the voltage value at regular intervals, such as recording a voltage value of 220V or 218V at a certain moment, which constitutes the voltage data segment.

[0018] In addition, temperature gradient data is collected by temperature sensors installed in key parts of the equipment. The sensors monitor temperature changes in real time and calculate the temperature gradient based on their positional relationships. For example, sensors are placed at different locations on the outer shell of a transformer to collect the temperature at different locations and then calculate the temperature gradient data.

[0019] Optionally, the meteorological data comes from surrounding meteorological monitoring stations and includes information such as temperature, humidity, and wind speed. This information also has a timestamp, such as recording a temperature of 25℃, humidity of 60%, and wind speed of 3m / s at a certain moment.

[0020] The aforementioned current time series data, voltage data segments, temperature gradient data, and meteorological correlation data together constitute a multi-source heterogeneous operation and maintenance data set.

[0021] S120: Perform cross-modal cleaning and time alignment processing on the operation and maintenance data set to generate a preprocessed data set.

[0022] In this embodiment of the invention, to ensure the accuracy and availability of the data, the collected operation and maintenance data set needs to undergo cross-modal cleaning and time alignment processing. Optionally, the cross-modal cleaning and time alignment processing of the operation and maintenance data set to generate a preprocessed data set includes:

[0023] S121: Divide the current time series data of the operation and maintenance data set into initial data segments according to a preset time granularity, and perform linear interpolation based on the adjacent data trend to generate interpolated current data segments for the missing initial data segments.

[0024] Taking the aforementioned large substation as an example, the preset time granularity is set to every minute. The collected current time-series data is divided into initial data segments by minute. If current data is missing for a certain minute, such as the 10th minute, linear interpolation is performed based on the trend of adjacent data. For example, if the current value is 50A in the 9th minute and 52A in the 11th minute, since current changes usually have a certain continuity, it can be considered that the current changes linearly within these two minutes. Then, through linear interpolation, the current value in the 10th minute can be calculated as (50+52)÷2=51A, thus generating the interpolated current data segment, ensuring the integrity of the current data.

[0025] S122: Detect abnormal voltage points in the voltage data segments of the operation and maintenance data set that exceed a preset safety threshold, and replace the abnormal voltage points with the historical average voltage of the corresponding devices to generate corrected voltage data segments.

[0026] For the voltage data segment of this substation, the preset safety threshold is set at 210V-230V. If a voltage value of 240V is detected at a certain moment in the collected voltage data, exceeding the preset safety threshold, it is considered an abnormal voltage point. In this case, the voltage data of the equipment over the past month is queried, and its historical average voltage is calculated. For example, if the historical average voltage is 225V, then the abnormal voltage point of 240V is replaced with 225V, thereby generating a corrected voltage data segment that better reflects the normal operating condition of the equipment.

[0027] S123: Perform sliding window mean filtering on the temperature gradient data of the operation and maintenance data set, identify abnormal fluctuation intervals where the temperature change rate exceeds the threshold within the window, and use Gaussian filtering to generate denoised temperature data segments.

[0028] Taking the temperature gradient data of the transformer in this substation as an example, the sliding window size is set to 10 time points, meaning each window consists of 10 consecutive temperature gradient data points. The mean is calculated within each window. For example, if the temperature gradient data within a window are 0.5℃ / m, 0.6℃ / m, 0.4℃ / m, etc., the mean is calculated as (0.5 + 0.6 + 0.4) ÷ 3 = 0.5℃ / m. Then, the rate of temperature change within the window is calculated. If the rate of temperature change within a window exceeds a preset threshold of 0.2℃ / m / min, that window is identified as an abnormal fluctuation range. For these abnormal fluctuation ranges, Gaussian filtering is used for noise reduction. Gaussian filtering, based on the distribution characteristics of the Gaussian function, performs weighted averaging on the data, making the abnormal fluctuation data closer to the normal fluctuation range, thus generating denoised temperature data segments.

[0029] S124: Extract the timestamp information of the meteorological data associated with the operation and maintenance data set, and synchronize and align the timestamp information of the meteorological data associated with the timestamps of the interpolated current data segment, the corrected voltage data segment, and the noise-reduced temperature data segment to generate a synchronized environmental impact factor sequence.

[0030] It's understandable that meteorological data contains timestamps. For example, if a meteorological monitoring data point is recorded at 10:00:00, current, voltage, and temperature data also have their own corresponding timestamps. The timestamps of the meteorological data are compared with those of the interpolated current data segment, the corrected voltage data segment, and the noise-reduced temperature data segment. If the current data was recorded at 10:00:05, the voltage data at 10:00:03, and the temperature data at 10:00:07, the timestamps of these data are synchronized by adjusting or selecting the closest time point. For example, all data can be aligned to 10:00:05. The synchronized meteorological data (temperature, humidity, wind speed, etc.) is then combined with the current, voltage, and temperature data to form a synchronized environmental impact factor sequence, ensuring temporal consistency across different data types.

[0031] S125: The preprocessed data set is generated by uniformly encoding the interpolated current data segment, the corrected voltage data segment, the noise-reduced temperature data segment, and the synchronized environmental impact factor sequence.

[0032] In this step, a unified time encoding method is used for the interpolated current data segment, the corrected voltage data segment, the noise-reduced temperature data segment, and the synchronized environmental impact factor sequence. For example, the time corresponding to each data point is converted into the number of seconds calculated from a fixed start time (e.g., 00:00:00 of the current day). If the time of a data point is recorded as 10:01:00, the number of seconds is 10 × 3600 + 1 × 60 + 0 = 36060 seconds. All data is time encoded in the above way, and all encoded data are integrated to generate a preprocessed dataset, providing a unified and standardized data foundation for subsequent analysis and processing.

[0033] S130: Perform dynamic feature fusion and correlation analysis on the preprocessed data set to generate a device status feature set and a device-environment interaction feature set.

[0034] In this embodiment of the invention, to gain a deeper understanding of the device status and the interaction between the device and the environment, dynamic feature fusion and correlation analysis need to be performed on the preprocessed data set. Optionally, performing dynamic feature fusion and correlation analysis on the preprocessed data set to generate a device status feature set and a device-environment interaction feature set includes:

[0035] S131: Perform multi-scale wavelet decomposition on the interpolated current data segment to generate fluctuation components and trend components, and calculate the energy density ratio of each component.

[0036] For the interpolated current data segment, taking the previously generated current data segment as an example, a multi-scale wavelet decomposition method is used for processing. Wavelet decomposition can decompose the signal into components of different frequencies. In this embodiment of the invention, the current data is decomposed into fluctuation components and trend components. For example, after decomposing the current data, the fluctuation component reflects the short-term rapid changes in the current, while the trend component reflects the long-term trend of the current. Subsequently, the energy density of the fluctuation component is calculated. For example, if the value of the fluctuation component in a time period is 5A, 6A, 4A, etc., the energy density is calculated by adding the squares of each value and then dividing by the length of the time period, such as (5² + 6² + 4A). 2 )÷3. Similarly, calculate the energy density of the trend component, and then calculate the ratio of the energy density of the fluctuation component to the energy density of the trend component. This ratio serves as a characteristic value, reflecting the relative relationship between fluctuation and trend in the current data.

[0037] S132: Perform density clustering on the corrected voltage data segment to generate the distribution density of abnormal voltage points, and calculate the regional cumulative anomaly index based on the clustering results.

[0038] For example, clustering parameters can be set to classify voltage data according to their distribution density. In one region, voltage data might be concentrated around 220V, while in another region it might be concentrated around 215V. Density clustering can separate the voltage data from these different regions. The number of abnormal voltage points (voltage points outside the normal range) within each cluster is calculated, and then divided by the total number of voltage data points in that region to obtain the abnormal voltage point distribution density. For example, if a cluster has 100 voltage data points, and 10 of them are abnormal, the abnormal voltage point distribution density is 10 ÷ 100 = 0.1. Based on the abnormal voltage point distribution density of each cluster, the regional cumulative anomaly index is calculated. For example, if this substation has three clusters with abnormal voltage point distribution densities of 0.1, 0.05, and 0.15 respectively, the regional cumulative anomaly index is calculated as (0.1 + 0.05 + 0.15) ÷ 3 = 0.1. This index reflects the overall degree of voltage anomalies within the entire region.

[0039] S133: Perform slope analysis on the noise-reduced temperature data segment to extract the temperature rise rate, and count the number of times the temperature fluctuation exceeds the threshold per unit time to generate the abnormal fluctuation frequency.

[0040] Optionally, the denoised temperature data segment can be analyzed. Taking transformer temperature data as an example, the temperature difference between adjacent time points can be calculated and then divided by the time interval to obtain the temperature rise rate. For example, if the temperature rises from 30℃ to 35℃ within 10 minutes (600 seconds), the temperature rise rate is (35-30)÷600=0.0083℃ / s. A temperature fluctuation threshold of 1℃ is set, and the number of times the temperature fluctuates by more than 1℃ within a unit of time (e.g., one hour) is counted. For example, if the temperature fluctuates by more than 1℃ 5 times within one hour, the abnormal fluctuation frequency is 5 times / hour. These two characteristic values ​​reflect the temperature rise trend and fluctuation situation, respectively.

[0041] S134: Input the synchronized environmental impact factor sequence into the pre-trained environmental weight prediction model and output a combined vector of temperature impact weight, humidity impact weight and wind speed impact weight.

[0042] Optionally, the synchronized sequence of environmental impact factors (including information such as temperature, humidity, and wind speed) can be input into a pre-trained environmental weight prediction model. This model is trained using a large amount of historical data and can predict the impact weights of temperature, humidity, and wind speed on equipment operation based on the input environmental data. For example, based on the current input environmental data, the model outputs a weight of 0.6 for temperature, 0.2 for humidity, and 0.2 for wind speed. These three weights form a combined vector [0.6, 0.2, 0.2], which represents the degree of influence of environmental factors on the equipment status.

[0043] S135: The energy density ratio, the abnormal voltage point distribution density, the regional cumulative anomaly index, the temperature rise rate, and the combined vector are concatenated to generate a device state feature set and a device-environment interaction feature set; wherein, the energy density ratio, the regional cumulative anomaly index, and the temperature rise rate have been subjected to max-min normalization processing before feature concatenation.

[0044] First, the energy density ratio, regional cumulative anomaly index, and temperature rise rate are normalized using a maximum-minimum normalization process. Taking the energy density ratio as an example, its value ranges from 0.1 to 0.9. The normalization formula (x-min) ÷ (max-min) maps it to the range of 0 to 1. For example, if an energy density ratio value is 0.5, the normalized value is (0.5-0.1) ÷ (0.9-0.1) = 0.5. The regional cumulative anomaly index and temperature rise rate are processed in the same way. Then, the normalized energy density ratio, abnormal voltage point distribution density, regional cumulative anomaly index, temperature rise rate, and combined vector are concatenated. For example, these features are arranged in a certain order to form a multi-dimensional feature vector, such as [0.5, 0.1, 0.1, 0.0083, 0.6, 0.2, 0.2]. This feature vector constitutes part of the equipment status feature set and the equipment-environment interaction feature set. In this way, different types of features are integrated, providing more comprehensive data support for subsequent visualization and analysis.

[0045] S140: Based on the device state feature set and the device-environment interaction feature set, generate an interactive visualization element set through a multi-dimensional visualization strategy.

[0046] In this embodiment of the invention, to more intuitively display the device status and the interaction between the device and the environment, an interactive visualization element set is generated based on the aforementioned generated feature set using a multi-dimensional visualization strategy. Optionally, generating the interactive visualization element set based on the device status feature set and the device-environment interaction feature set using a multi-dimensional visualization strategy includes:

[0047] S141: A multi-view projection strategy is used to map the energy density ratio and the regional cumulative anomaly index to a three-dimensional visualization space, wherein node size parameters of a dynamic topology graph are generated based on the energy density ratio, and node color wheel gradient parameters are generated based on the regional cumulative anomaly index.

[0048] In the 3D visualization space, for energy density ratios, regions with higher energy density ratios correspond to larger dynamic topology graph node sizes, while regions with lower energy density ratios have smaller node sizes. For example, if the energy density ratio ranges from 0 to 1, an energy density ratio of 0.8 is mapped to a larger node size, such as a radius of 5 units; while an energy density ratio of 0.2 is mapped to a smaller node size, such as a radius of 1 unit. For the regional cumulative anomaly index, node color wheel gradient parameters are generated based on its value. If the regional cumulative anomaly index is high, the color wheel leans towards red; if it is low, the color wheel leans towards green. For example, when the regional cumulative anomaly index is 0.8, the color wheel is dark red; when the regional cumulative anomaly index is 0.2, the color wheel is light green. In this way, the energy density ratio and regional cumulative anomaly index are displayed in a visually intuitive graphical way in the 3D visualization space.

[0049] S142: Based on the layered mapping strategy, the temperature rise rate is generated into the color block elevation parameters of the three-dimensional heat map through height field mapping, and the color block pulsation parameters are generated based on the abnormal fluctuation frequency through the transparency decay function.

[0050] Regarding the rate of temperature rise, taking a 3D heatmap of a certain area as an example, the higher the rate of temperature rise, the higher the elevation of the corresponding color block. For example, when the rate of temperature rise is 0.01℃ / s, the elevation of the color block is set to 10 units; when the rate of temperature rise is 0.005℃ / s, the elevation of the color block is set to 5 units. For the frequency of abnormal fluctuations, the color block pulsation parameters are generated using a transparency decay function. For example, when the frequency of abnormal fluctuations is high, the color block transparency is low and the pulsation speed is fast; when the frequency of abnormal fluctuations is low, the color block transparency is high and the pulsation speed is slow. For example, if the frequency of abnormal fluctuations is 10 times / hour, the transparency of the color block is calculated to be 0.3, and the pulsation speed is 3 jumps per second; if the frequency of abnormal fluctuations is 2 times / hour, the transparency is 0.8, and the pulsation speed is 1 jump per second. In this way, through height field mapping and the transparency decay function, the rate of temperature rise and the frequency of abnormal fluctuations are displayed in the 3D heatmap with different visual characteristics.

[0051] S143: Using a multi-dimensional vector fusion algorithm, the temperature influence weight, humidity influence weight, and wind speed influence weight in the combined vector are mapped to the arrow length parameter, feathering parameter, and tilt angle parameter of the meteorological vector, respectively.

[0052] Optionally, a multi-dimensional vector fusion algorithm is used to map the influence weights of temperature, humidity, and wind speed in the combined vector. For example, the higher the temperature influence weight, the longer the arrow length of the meteorological vector. For instance, when the temperature influence weight is 0.8, the arrow length is 10 units; when the temperature influence weight is 0.2, the arrow length is 2 units. The humidity influence weight corresponds to a feathering parameter; the higher the humidity influence weight, the more pronounced the feathering effect. For example, when the humidity influence weight is 0.6, the feathering parameter is set to give the arrow edges a wider blurring effect; when the humidity influence weight is 0.2, the feathering effect is weaker. The wind speed influence weight corresponds to an inclination angle parameter; the higher the wind speed influence weight, the larger the arrow inclination angle. For example, when the wind speed influence weight is 0.7, the arrow inclination angle is 60°; when the wind speed influence weight is 0.3, the arrow inclination angle is 30°. In this way, the influence weights of environmental factors are intuitively displayed in the form of meteorological vectors.

[0053] S144: Establish a cross-view association mechanism to synchronize the spatial coordinates of the node size parameters and node color wheel gradient parameters of the dynamic topology map, the color block elevation parameters and color block pulsation parameters of the three-dimensional heat map, and the arrow length parameters, feathering parameters and tilt angle parameters of the meteorological vector, thereby generating an interactive visualization element set with coordinate linkage.

[0054] Optionally, a cross-view association mechanism can be established to ensure that elements in different views can be interconnected and interact. For example, when a node in a dynamic topology map changes its size or color ring gradient parameters due to changes in energy density ratio or regional cumulative anomaly index, the elevation parameters and pulsation parameters of the corresponding area in the 3D heatmap will also be adjusted accordingly through spatial coordinate synchronization. The parameters of the meteorological vector will also change synchronously according to the comprehensive influence of environmental factors. For example, if a node in the dynamic topology map increases in size due to an increase in energy density ratio, the elevation parameters of the corresponding area in the associated 3D heatmap will also increase, and the arrow length of the meteorological vector may lengthen due to changes in the weight of temperature influence, forming an interactive visual element set with coordinate linkage. Users can gain a deeper understanding of the interaction between equipment status and environmental factors by operating and observing the linkage between these elements.

[0055] S150: Dynamically render and optimize the layout of the interactive visualization element set, and output a real-time updated power grid operation and maintenance visualization interface; wherein, the power grid operation and maintenance visualization interface is configured with a user operation priority response mechanism for handling interaction conflicts.

[0056] In this embodiment of the invention, to provide users with a clear, smooth, and user-friendly visual interface, it is necessary to dynamically render and optimize the layout of the interactive visual element set, and configure a user operation priority response mechanism. Optionally, the dynamic rendering and layout optimization of the interactive visual element set to output a real-time updated power grid operation and maintenance visual interface includes:

[0057] S151: Based on the node size parameters and node color wheel gradient parameters of the dynamic topology graph in the interactive visualization element set, create a topology view in a three-dimensional spatial projection coordinate system.

[0058] In this embodiment of the invention, a topology view showing the connection relationship of power grid equipment is constructed as an example, and a power grid area containing multiple substations and transmission lines is taken as an example. Each substation is represented by a node in the dynamic topology diagram. According to the previously determined node size parameters, for example, an important substation with a high energy density ratio has a corresponding node size parameter set to a radius of 10 units, while some small substations may have a node radius of 3 units. For the node color ring gradient parameter, if the cumulative anomaly index of a substation area is high, the color ring is displayed as dark red, such as the RGB value of the color ring being (255, 0, 0); if the anomaly index is low, the color ring is light green, with an RGB value of (0, 255, 0). In the three-dimensional spatial projection coordinate system, these nodes are arranged according to their actual connection relationship in the power grid. Transmission lines are connected to each node with lines. The thickness of the lines can be set according to factors such as the current carrying capacity of the transmission lines. For example, lines with high current carrying capacity have thicker lines, with a width of 5 units, while those with low current carrying capacity have a width of 2 units. This creates a topological view in a three-dimensional spatial projection coordinate system that can intuitively display the connection relationships of power grid equipment and the relevant characteristics of each equipment.

[0059] S152: Based on the elevation parameters and pulsation parameters of the color blocks in the three-dimensional heat map, a heat map with a height field gradient is generated by superimposing it in the three-dimensional spatial projection coordinate system.

[0060] Taking a 3D heatmap of the transformer area as an example, a heatmap is generated by overlaying previously determined color block elevation and pulsation parameters. For transformers in different locations, if a transformer has a faster temperature rise rate, its corresponding color block elevation parameter is higher, for example, 15 units of height, while the color block elevation of a transformer with a slower temperature rise rate is 5 units of height. Regarding the color block pulsation parameter, if a transformer has a high frequency of abnormal fluctuations (4 pulsations per second), its transparency is 0.4; if a transformer has a low frequency of abnormal fluctuations (1 pulsation per second), its transparency is 0.8. In a 3D spatial projection coordinate system, these color blocks with different elevation and pulsation characteristics are overlaid according to the actual location of the transformer, forming a heatmap with a height field gradient. From this view, the temperature changes of transformers in different locations can be seen intuitively; high-elevation areas indicate rapid temperature rise, and areas with obvious pulsation indicate large temperature fluctuations, facilitating maintenance personnel to quickly locate potentially problematic equipment areas.

[0061] S153: Map the arrow length parameter, feathering parameter, and tilt angle parameter of the meteorological vector to the three-dimensional spatial projection coordinate system to generate a dynamic vector field meteorological view.

[0062] Taking the meteorological conditions of the power grid area as an example, the relevant parameters of meteorological vectors are mapped onto a three-dimensional spatial projection coordinate system to generate a meteorological view. For areas with a high weighting for temperature influence, such as a weighting of 0.7, the corresponding meteorological vector arrow length is set to 12 units. When the weighting for humidity influence is 0.5, a feathering parameter gives the arrow edges a moderate degree of blurring. When the weighting for wind speed influence is 0.6, the arrow tilt angle is 50°. In the three-dimensional spatial projection coordinate system, meteorological vectors with different lengths, feathering, and tilt angles are drawn according to meteorological data from different locations. Each meteorological vector represents the comprehensive meteorological influence at a specific location, and multiple meteorological vectors form a dynamic vector field meteorological view. Users can use this view to understand the direction and degree of influence of meteorological factors on power grid equipment in different areas; for example, a longer arrow indicates a greater temperature influence, and a larger tilt angle indicates a greater wind speed influence.

[0063] S154: Construct an adaptive layout matrix based on the user terminal window size, arrange the topology view, thermal view and meteorological view in space according to a preset division ratio, and establish the coordinate mapping relationship between the three views.

[0064] In this embodiment of the invention, the user terminal window size is 1920×1080 pixels. First, an adaptive layout matrix is ​​constructed based on preset segmentation ratios, for example, the topology view occupies 40% of the window width and 50% of the height; the thermal view occupies 30% of the window width and 40% of the height; and the meteorological view occupies 30% of the window width and 40% of the height. In this layout, the topology view is placed in the left-hand portion of the window, the thermal view in the upper right corner, and the meteorological view in the lower right corner. Then, a coordinate mapping relationship is established between the three views. For example, if the coordinates of a node in the topology view are (x1, y1, z1), the corresponding relative position coordinates (x2, y2, z2) and (x3, y3, z3) in the thermal and meteorological views can be determined through the coordinate mapping relationship. Thus, when an element in one view changes, corresponding interactive displays can be made in other views based on the coordinate mapping relationship, facilitating users to comprehensively view information from different aspects.

[0065] S155: When a focusing operation is detected in a node region in the topology view, the elevation parameters of the color blocks corresponding to the spatial coordinates in the thermal view are adjusted synchronously based on the coordinate mapping relationship, and the transparency of the feathering parameters of the related region vectors in the meteorological view is increased.

[0066] For example, when a user focuses on a substation node in the topology view, the system detects this action. Based on the previously established coordinate mapping, it locates the corresponding color block in the thermal view. For instance, if the elevation parameter of the color block corresponding to this substation in the thermal view was originally 8 units, it is increased to 12 units to highlight the temperature changes in that area. Simultaneously, the system finds the vector of the related area in the meteorological view and increases its feathering parameter transparency from 0.5 to 0.8. In this way, when a user focuses on a device in the topology view, they can simultaneously and intuitively see the enhanced temperature changes in the area where that device is located in the thermal view and a clearer display of the influence of related meteorological factors in the meteorological view, providing a more comprehensive and interconnected information display.

[0067] S156: Perform spatiotemporal interpolation calculation on the motion trajectory of the meteorological vector, generate trajectory smoothing parameters based on the real-time change rate of wind speed influence weight, and use a Bezier curve fitting algorithm for dynamic rendering.

[0068] It's understandable that meteorological vectors continuously change their position and parameters over time. First, spatiotemporal interpolation is performed on the meteorological vector's trajectory. For example, if a meteorological vector moves from position A to position B within a certain time period, spatiotemporal interpolation can yield the approximate position of the vector at the intermediate moment. Based on the real-time change rate of the wind speed influence weight (e.g., if the wind speed influence weight changes from 0.4 to 0.6 within a certain time period, the change rate is 0.2 / unit of time), trajectory smoothing parameters are generated. Then, a Bézier curve fitting algorithm is used to dynamically render the meteorological vector's trajectory. Bézier curves make the vector's trajectory smoother and more natural. For example, by fitting the starting point, control point, and ending point of the meteorological vector using a Bézier curve, the vector presents a smooth curved trajectory during movement, rather than a rigid straight line, improving the visualization effect and user experience.

[0069] S157: Construct the abnormal fluctuation propagation path in the thermal view, calculate the path confidence interval based on the temperature rise rate and historical propagation pattern, and generate a gradient color light band through transparency gradient mapping.

[0070] In the thermal view, an abnormal temperature fluctuation in a transformer is used as an example to construct the propagation path of the abnormal fluctuation. Based on the transformer's temperature rise rate, for example, 0.015℃ / s, and combining historical propagation patterns of abnormal fluctuations under similar temperature rise rates, possible propagation paths of the abnormal fluctuations are calculated through data analysis and modeling. For example, according to historical data, when the temperature rise rate is between 0.01-0.02℃ / s, abnormal fluctuations typically propagate along specific lines to surrounding equipment. The confidence interval for this propagation path is calculated, for example, 80%. Then, a gradient color light band is generated through transparency gradient mapping. For the high confidence portion, the light band has lower transparency and a darker color, such as red at a transparency of 0.3; for the low confidence portion, the light band has higher transparency and a lighter color, such as light red at a transparency of 0.8. In this way, the possible propagation direction and confidence level of the abnormal fluctuation can be visually observed in the thermal view.

[0071] S158: Real-time detection of the geometric collision relationships of each element in the dynamic topology map, the three-dimensional heat map, and the meteorological view. When the proportion of the volume of the intersection of the node bounding box of the dynamic topology map and the color block bounding box of the three-dimensional heat map to the volume of the smaller bounding box exceeds a preset percentage threshold, the spherical unfolding algorithm is activated to decouple the collision view by azimuth angle and reorganize the space.

[0072] For example, during real-time detection, the bounding box of a node in the dynamic topology graph is set to a sphere with a radius of 8 units, and the bounding box of a color block in the 3D heatmap is set to a cuboid with length, width, and height of 10, 8, and 6 units respectively. Their intersection volume is calculated; for example, if the intersection volume is 200 cubic units, and the smaller bounding box (cube) has a volume of 480 cubic units, the intersection volume ratio is 200 ÷ 480 ≈ 41.7%. When this ratio exceeds a preset percentage threshold, such as 30%, the spherical unfolding algorithm is activated. This algorithm first decouples the nodes and color blocks from their azimuth angles, displaying them separately at different angles. Then, it reorganizes the space, for example, rearranging the nodes and color blocks in space to avoid visual overlap and confusion, allowing users to clearly see each element without affecting information viewing and analysis.

[0073] In an optional embodiment, the method further includes:

[0074] S210: Monitor the user's touch operation trajectory on the power grid operation and maintenance visualization interface, and extract the acceleration characteristics and contact area change rate of the touch operation trajectory.

[0075] When a user interacts with the power grid operation and maintenance visualization interface using a touchscreen, the system begins monitoring the user's touch operation trajectory. For example, when a user slides their finger across the screen to view information in different areas, the system records the finger's position coordinates on the screen at each moment, thus forming a touch operation trajectory. For this trajectory, its acceleration characteristics are calculated. For instance, over a period of time, the finger moves from position (x1, y1) to (x2, y2) over a time interval t. The acceleration is obtained by calculating the ratio of the change in velocity to time. If the initial velocity is v1, and after time t the velocity becomes v2, the acceleration a = (v2 - v1) / t. Simultaneously, the rate of change of contact area is calculated. The contact area when the finger first touches the screen is S1, and as the finger slides, the contact area changes, becoming S2 at a certain moment over a time interval Δt. The rate of change of contact area = (S2 - S1) / Δt. These characteristics will be used for subsequent operation judgment and processing.

[0076] S220: Match a preset view zoom instruction set according to the acceleration characteristics, and generate hot zone focusing weight parameters according to the contact area change rate.

[0077] The preset view scaling instruction set includes scaling operations corresponding to different acceleration ranges. For example, when the acceleration a is greater than a certain threshold a1, it corresponds to a view magnification operation; when the acceleration a is less than a certain threshold a2 (a2 < a1), it corresponds to a view reduction operation. For example, when the acceleration a = 0.5 (the unit is set according to the actual situation), by matching with the preset instruction set, it is determined as a view magnification operation. For the contact area change rate, when the contact area change rate is positive and large, it indicates that the user may be more concerned about the current contact area, and a higher hot zone focus weight parameter is generated. For example, when the contact area change rate is 0.2 (the unit is set according to the actual situation), the hot zone focus weight parameter generated according to the preset rule is 0.8, indicating that this area has a higher focus weight, and subsequent operations will pay more attention to this area.

[0078] S230: Based on the hot zone focus weight parameter, adjust the scaling ratio of the node size parameter of the dynamic topology map, and simultaneously increase the blinking frequency of the associated color block pulsation parameter in the three-dimensional heat map.

[0079] According to the generated hot zone focus weight parameter, which is 0.8, for the node size parameter of the dynamic topology map located in this hot zone, increase its scaling ratio. For example, the original node radius was 5 units, and now it is scaled to 7 units to make it more prominent in the topology map. At the same time, for the color block pulsation parameter associated with this hot zone in the three-dimensional heat map, increase its blinking frequency. For example, the original color block pulsated 2 times per second, and now it is increased to 4 times per second. This can highlight the hot zone that the user is concerned about in different views, facilitating the user to more clearly view the device status information in this area.

[0080] S240: Input the adjusted node size parameter and color block pulsation parameter into the particle rendering module to generate a touch feedback particle flow, and superimpose it on the edge area of the power grid operation and maintenance visualization interface; among them, the scaling instruction of single-point touch is preferentially responded to, and when multi-point touch is detected, an operation type conflict arbitration mechanism is enabled.

[0081] The adjusted node size and color block pulsation parameters are input into the particle rendering module. Based on these parameters, the module generates a touch feedback particle stream. For example, as the node size increases and the color block pulsation frequency rises, more, denser, and more noticeably flickering particles are generated. These particle streams are superimposed onto the edge area of ​​the power grid operation and maintenance visualization interface, creating a visual feedback effect that lets the user know their actions have been responded to by the system. Regarding touch operation processing, single-point touch zoom commands are prioritized. When multi-touch is detected, such as when a user performs different operations with two fingers simultaneously, an operation type conflict arbitration mechanism is activated. This mechanism, based on preset rules such as first determining the priority of operations or merging operations based on their similarity, ensures that the system can correctly handle complex multi-touch operations and avoids display anomalies or functional confusion caused by operation conflicts.

[0082] In an optional embodiment, the method further includes:

[0083] S310: Real-time analysis of the regional cumulative anomaly index in the set of device status features; when the index exceeds a preset risk threshold, activate the anomaly tracing mode.

[0084] The system continuously analyzes the regional cumulative anomaly index from the device status feature set in real time. For example, the preset risk threshold is set to 0.6. If the calculated regional cumulative anomaly index for a region is 0.7, exceeding the preset risk threshold, the system automatically activates the anomaly tracing mode. In this mode, the system will conduct an in-depth analysis of the causes and possible propagation paths of the device anomalies in that region, enabling maintenance personnel to take timely measures to prevent the anomalies from escalating further.

[0085] S320: Extract the correlation matrix between the frequency of abnormal fluctuations and the rate of temperature rise of the same equipment in the historical operation and maintenance data set under the anomaly tracing mode.

[0086] After activating the anomaly tracing mode, the system searches for data on the same equipment in the historical operation and maintenance data set. For example, for a specific model of transformer, it extracts its operation and maintenance data from the past year and analyzes the relationship between the frequency of abnormal fluctuations and the rate of temperature rise. By statistically analyzing this data, a correlation matrix is ​​constructed. For example, in different time periods, the frequency of abnormal fluctuations is 5 times / month, 3 times / month, etc., and the rate of temperature rise is 0.005℃ / s, 0.01℃ / s, etc., respectively. This data is then organized into a matrix (i.e., a correlation matrix). This correlation matrix can help analyze the potential relationship between abnormal fluctuations and the rate of temperature rise, providing a basis for subsequently determining the anomaly propagation path.

[0087] S330: Generate a probability density distribution map of the abnormal propagation path of the device based on the correlation matrix, and encode the probability density distribution map as a semi-transparent thermal overlay layer; wherein, if historical data is missing, the average correlation matrix of similar devices is used as the default value.

[0088] Optionally, based on the constructed correlation matrix, a probability density distribution map of the equipment anomaly propagation path is generated using data analysis and modeling algorithms. For example, by analyzing the data in the correlation matrix, it is found that when the temperature rise rate reaches a certain value, abnormal fluctuations are more likely to propagate along certain specific lines. These propagation paths and their corresponding probability densities are then graphically displayed to form a probability density distribution map. This distribution map is then encoded as a semi-transparent thermal overlay layer, allowing it to be superimposed on a three-dimensional thermal map. If historical data is missing and a correlation matrix for the specific equipment cannot be constructed, the average correlation matrix of similar equipment is used as the default value. For example, for transformers of the same model produced by the same manufacturer, the average correlation matrix data of other similar transformers is used for subsequent analysis and processing to ensure that the anomaly tracing process can continue.

[0089] S340: The semi-transparent thermal overlay is embedded in the region below the elevation parameters of the color blocks in the 3D heatmap using a depth buffer layered rendering strategy.

[0090] Optionally, a depth buffer layered rendering strategy is used to embed the encoded semi-transparent thermal overlay into the 3D heatmap. The depth buffer records the depth information of each pixel in 3D space, which is used to determine the display order of different layers. In this embodiment, the semi-transparent thermal overlay is placed below the elevation parameters of the color blocks in the 3D heatmap. For example, when rendering the 3D heatmap, the elevation information of the color blocks is rendered first, and then, based on the information from the depth buffer, the semi-transparent thermal overlay is accurately drawn below the elevation of the color blocks. This allows users to simultaneously see the device's temperature and the probability distribution of abnormal propagation paths when viewing the heatmap, providing a more intuitive information display for analyzing device anomalies.

[0091] In an optional embodiment, the method further includes:

[0092] S410: Continuously captures time-series changes in the arrow length and tilt angle parameters of meteorological vectors, and generates a dynamic decay curve of the wind speed influence weight.

[0093] The system continuously monitors the changes in the arrow length and tilt angle parameters of meteorological vectors over time. For example, these two parameters are recorded every minute. Or, for instance, within an hour, the arrow length gradually decreases from 10 units to 8 units, and the tilt angle changes from 45° to 30°. Based on this recorded data, a dynamic decay curve of the wind speed impact weight is generated, with time on the horizontal axis and the wind speed impact weight (calculated from the arrow length and tilt angle parameters using a specific algorithm, e.g., wind speed impact weight = arrow length × sin(tilt angle)) on the vertical axis. This curve visually demonstrates the changing trend of the wind speed impact weight over time, helping maintenance personnel understand the dynamic process of meteorological factors affecting power grid equipment.

[0094] S420: Perform extreme point detection on the dynamic decay curve, extract abnormal time segments with sudden slope changes in the curve, and mark them as meteorological event trigger points.

[0095] The generated dynamic decay curves are analyzed to detect extreme points. For example, mathematical methods such as differentiation are used to find points where the curve slope is 0; these points may be maxima or minima. Simultaneously, attention is paid to sections of the curve where the slope abruptly changes, i.e., where the curve change suddenly intensifies or slows down. For example, if at a certain moment the curve slope suddenly changes from 0.2 to -0.5, this time segment is identified as an anomalous time segment and marked as a meteorological event trigger point. These trigger points may correspond to sudden changes in meteorological conditions, potentially impacting the operation of power grid equipment and requiring further monitoring and analysis.

[0096] S430: Based on the meteorological event trigger point, retrieve the temperature influence weight at the corresponding moment in the synchronized environmental impact factor sequence.

[0097] Once the meteorological event trigger point is identified, the temperature impact weight for that time is retrieved from the synchronized environmental impact factor sequence based on the corresponding time. For example, if the meteorological event trigger point is 14:30, the temperature impact weight recorded for that time in the environmental impact factor sequence is 0.5. By obtaining this temperature impact weight, we can understand the degree to which temperature affects equipment operation when meteorological conditions change, providing data support for subsequent analysis of the comprehensive impact of meteorological events on equipment.

[0098] S440: Normalize and fuse the temperature influence weight and wind speed influence weight to generate a meteorological event alarm sign and insert it into the end of the arrow in the dynamic vector field meteorological view; wherein, a time tolerance window is set to compensate for data delay.

[0099] In practice, the first step is to normalize the weights for temperature and wind speed. Normalization adjusts these two weights to a uniform range, facilitating subsequent fusion calculations.

[0100] To determine the temperature influence weight, first define its range. For example, in historical data analysis, the minimum temperature influence weight is 0.1, and the maximum is 0.9. For a specific temperature influence weight value, say 0.6, the normalization process is as follows: subtract the minimum value from this value, then divide by the difference between the maximum and minimum values. That is, (0.6 - 0.1) ÷ (0.9 - 0.1) = 0.5 ÷ 0.8 = 0.625, thus obtaining the normalized temperature influence weight.

[0101] Regarding the weighting of wind speed influence, we first need to define its value range. We set the minimum wind speed influence weight to 0.2 and the maximum to 0.8, with the current wind speed influence weight being 0.5. The normalized calculation is then (0.5-0.2)÷(0.8-0.2)=0.3÷0.6=0.5, which is the normalized wind speed influence weight.

[0102] Next, a fusion calculation is performed using a simple weighted average method. In this embodiment of the invention, a fusion coefficient is set for the temperature influence weight, which is 0.6. Therefore, the fusion coefficient for the wind speed influence weight is 1 - 0.6 = 0.4. The fused weight is calculated as follows: multiply the normalized temperature influence weight by the fusion coefficient of the temperature influence weight, and then add the normalized wind speed influence weight multiplied by the fusion coefficient of the wind speed influence weight. That is, the fused weight = normalized temperature influence weight × 0.6 + normalized wind speed influence weight × 0.4. Substituting the previously calculated normalized temperature influence weight of 0.625 and normalized wind speed influence weight of 0.5 into the equation, we get the fused weight = 0.625 × 0.6 + 0.5 × 0.4 = 0.375 + 0.2 = 0.575.

[0103] Weather event alarm icons are generated based on the fused weights. Pre-defined rules are used: a red alarm icon is generated when the fused weight value is greater than or equal to 0.7, indicating a high risk to the equipment; a yellow warning icon is generated when the weight value is between 0.4 and 0.7, indicating some risk requiring attention; and a green normal icon is generated when the weight value is less than 0.4. Since the calculated fused weight is 0.575, a yellow warning icon is generated.

[0104] During data processing, delays may occur in data acquisition and transmission. Therefore, a time tolerance window needs to be set to compensate for data latency. For example, the time tolerance window can be set to 5 minutes before and after the data acquisition and transmission. When a meteorological event trigger point is detected, the corresponding temperature impact weight is searched within the 5-minute time period before and after the trigger point for the aforementioned fusion calculation. This ensures that the generated alarm flag can more accurately reflect the actual impact of the meteorological event on the equipment.

[0105] Finally, the generated meteorological event alarm icons are inserted at the ends of the arrows in the dynamic vector field meteorological view. This allows users to visually see the meteorological event alarm status corresponding to each meteorological vector in the dynamic vector field meteorological view, enabling them to quickly understand the potential impact of meteorological factors on power grid equipment.

[0106] In an optional embodiment, the method further includes:

[0107] S510: Periodically read the node color wheel gradient parameters in the interactive visualization element set, and count the frequency of occurrence of each node color wheel gradient parameter within a preset time window.

[0108] In this embodiment of the invention, the system reads the node color wheel gradient parameters in the interactive visualization element set according to a preset period. This period can be determined according to actual needs, for example, it can be set to perform a reading operation once every 15 minutes.

[0109] Taking a dynamic topology graph with multiple nodes as an example, each node has different color wheel gradient parameters, which reflect certain state characteristics of the device represented by the node. A preset time window is set, such as one hour. Within this one hour, the system continuously records the occurrence of the color wheel gradient parameters of each node.

[0110] For example, the color wheel gradient parameter of node A exhibits a specific color and gradient effect, and the system detected this color wheel gradient parameter appearing 10 times within one hour; the color wheel gradient parameter of node B shows a different characteristic, appearing 8 times; the color wheel gradient parameter of node C is different again, appearing 12 times, and so on. Through the above statistical method, we can understand the frequency of change of the color wheel gradient parameter of each node during this period, providing important data for subsequent analysis of equipment status. This statistical data can reflect the stability and trend of equipment status. If the frequency of the color wheel gradient parameter of a certain node is abnormal, it may mean that the status of the equipment has changed significantly, requiring further attention and analysis.

[0111] S520: The occurrence frequency is processed by a preset device state degradation model, and the predicted device health value corresponding to different color ring gradients is output.

[0112] The preset equipment status degradation model is built on a large amount of historical operation and maintenance data and equipment operating characteristics. The function of this model is to predict the health status of the equipment based on the frequency of occurrence of node color wheel gradient parameters.

[0113] For each node's color wheel gradient parameter, the model performs a series of complex analyses and calculations. For example, if the color wheel gradient parameter of node A occurs 10 times, the model will first compare this frequency with the normal range of occurrences of the color wheel gradient parameter for that node in historical data. For instance, if historical data shows that the normal frequency range for the color wheel gradient parameter of that node is between 8 and 12 times, the model will then perform a comprehensive calculation based on this comparison, combined with other relevant factors such as device runtime and environmental factors.

[0114] The model may contain various computational logics and rules. For example, it might multiply the frequency of occurrence by different weighting coefficients based on its deviation from the normal range, and add the influence values ​​of other relevant factors to arrive at a final numerical value. This value, after further transformation and processing, yields the predicted device health value corresponding to the color wheel gradient. For instance, the model might calculate a predicted device health value of 0.7 for node A.

[0115] The color wheel gradient parameter for node B appears 8 times, and the model is processed in a similar manner as described above. By comparing with historical data and considering various relevant factors, the predicted device health value for node B is calculated to be 0.8.

[0116] The color wheel gradient parameter of node C appears 12 times. After a series of processing by the model, the corresponding predicted value of device health is 0.6.

[0117] These device health prediction values ​​are set between 0 and 1. Higher values ​​indicate better current device health, while lower values ​​suggest potential problems requiring close monitoring by maintenance personnel. By using a pre-defined device state degradation model to process the frequency of color wheel gradient parameters, this approach provides maintenance personnel with intuitive and valuable information about device health, helping to identify potential faults early and implement timely maintenance measures.

[0118] S530: Map the predicted equipment health value to the fill ratio parameter of the circular progress bar, and generate equipment health label according to the fill ratio parameter.

[0119] In this step, the predicted equipment health value is converted into a more intuitive visual representation, namely the fill ratio parameter of the circular progress bar, and the corresponding equipment health label is generated.

[0120] The mapping relationship between the predicted equipment health value and the fill ratio parameter of the circular progress bar is set to directly multiply the predicted equipment health value by 100%, and the result is the fill ratio of the circular progress bar. For example, if the predicted equipment health value for node A is 0.7, then its corresponding fill ratio parameter for the circular progress bar is 0.7 × 100% = 70%. This means that when the circular progress bar is displayed, it will be filled at a ratio of 70%.

[0121] The predicted health value of the device at node B is 0.8, and the corresponding fill ratio parameter of the circular progress bar is 0.8 × 100% = 80%.

[0122] The predicted health value of the device at node C is 0.6, and the corresponding fill ratio parameter of the circular progress bar is 0.6 × 100% = 60%.

[0123] Based on these fill ratio parameters, device health labels are generated. The label content must include both the fill ratio information and a corresponding description based on the health level represented by the fill ratio.

[0124] For node A, the fill ratio is 70%, and the generated device health label content is "Device health 70%, good condition".

[0125] For node B, the fill ratio is 80%, and the generated device health label content is "Device health 80%, good status".

[0126] For node C, the fill ratio is 60%, and the generated device health label content is "Device health 60%, needs attention".

[0127] By mapping the predicted health values ​​of devices to the fill ratio parameters of a circular progress bar and generating corresponding labels, maintenance personnel can understand the health status of each node at a glance when viewing the dynamic topology diagram, which facilitates quick location of devices that may have problems and improves maintenance efficiency.

[0128] S540: Dynamically bind the device health label to the outer region of the node in the dynamic topology graph, and update the edge transition state of the node color wheel gradient parameter through anti-aliased rendering; wherein, the Z-axis hierarchy relationship between the label and the node is processed through an independent rendering channel.

[0129] In a dynamic topology graph, the generated device health labels must be accurately bound to the corresponding outer area of ​​the node. This operation allows users to directly obtain the health information of the device represented by each node when viewing the topology graph.

[0130] For example, for node A, place the label "Device health 70%, good condition" in a suitable location around node A. The location should be chosen to avoid obstructing the node itself or other important information, while ensuring the user can clearly see the label content. For nodes B and C, follow the same method to accurately bind their respective device health labels to the outer area of ​​the corresponding node.

[0131] To enhance the user's visual experience and make the display of node color wheel gradients smoother and more natural, anti-aliasing rendering technology is used to update the edge transition states of the node color wheel gradient parameters. Anti-aliasing rendering can effectively reduce the jagged edges of the color wheel, making the color transitions more delicate and soft.

[0132] During rendering, the Z-axis hierarchy of labels and nodes is handled through independent rendering channels. This is because labels and nodes have different positional relationships in 3D space. Independent rendering channels allow for separate control over the display order and effect of labels and nodes along the Z-axis. For example, labels can be placed slightly above nodes, ensuring they are not obscured while still forming a cohesive visual effect. This way, when rendering dynamic topology diagrams, labels and nodes are presented to the user with a clear and logical hierarchy, further enhancing the readability and usability of the visualization interface and helping maintenance personnel better understand and analyze the status information of power grid equipment.

[0133] In a non-limiting embodiment, after outputting the real-time updated power grid operation and maintenance visualization interface, the method further includes: monitoring user multi-touch gestures on the power grid operation and maintenance visualization interface, extracting the radius of curvature and average contact pressure of the gesture trajectory; matching a preset view rotation sensitivity parameter based on the radius of curvature, and generating a depth-of-field offset for the touch focus area based on the average contact pressure; adjusting the Z-axis scaling factor of the node size parameters of the topology view in the three-dimensional spatial projection coordinate system based on the depth-of-field offset, and simultaneously reducing the elevation parameters of non-focus color blocks in the thermal view; inputting the adjusted Z-axis scaling factor and elevation parameters into a ray tracing engine to generate a visualization interface with dynamic depth-of-field effects; calculating the Euler angle change rate of the topology view in real time based on the view rotation sensitivity parameter, and smoothly rendering the viewpoint switching animation using a quaternion interpolation algorithm. In actual implementation, the X / Y axis ratio of the original node size parameters can be kept unchanged.

[0134] After outputting a real-time updated power grid operation and maintenance visualization interface, the system further enhances the user experience and visualization effects. When users perform multi-touch gestures on the power grid operation and maintenance visualization interface, the system responds quickly. For example, when users perform complex operations such as rotating and zooming on the screen with both hands, the system accurately extracts the radius of curvature of the gesture trajectory and the average contact pressure.

[0135] For example, a user's multi-touch gestures form a trajectory, and a specific algorithm calculates the radius of curvature of this trajectory. A smaller radius of curvature indicates a greater degree of bending in the gesture, potentially representing a more precise operation by the user; conversely, a larger radius of curvature indicates a smoother operation. Simultaneously, the system calculates the average contact pressure; higher contact pressure may indicate greater force applied by the user, suggesting a higher level of focus on the action.

[0136] Next, the system matches preset view rotation sensitivity parameters based on the radius of curvature. For example, a smaller radius of curvature results in a higher view rotation sensitivity parameter, meaning the view will rotate more sensitively to the user's gestures, satisfying the user's need for detailed viewing of specific areas. The system also generates a depth-of-field offset for the touch focus area based on the average contact pressure. A larger average contact pressure results in a correspondingly larger depth-of-field offset, making the focus area more prominent.

[0137] Based on the depth-of-field offset, the system adjusts the Z-axis scaling factor of the node size parameters in the topology view within the 3D spatial projection coordinate system. For example, when the depth-of-field offset increases, the Z-axis scaling factor is adjusted accordingly, making the nodes in the topology view appear more three-dimensional along the Z-axis. Simultaneously, the elevation parameters of non-focal color blocks in the heatmap view are reduced, highlighting information in the focal area and preventing excessive non-critical information from interfering with the user's view.

[0138] Next, the adjusted Z-axis scaling factor and altitude parameters are input into the ray tracing engine. The ray tracing engine simulates the propagation of light in the 3D scene based on these parameters, thereby generating a visualization interface with dynamic depth-of-field effects. In this interface, users can experience a more realistic and three-dimensional visual effect, as if observing the device in a real scene.

[0139] Finally, the system calculates the Euler angle change rate of the topology view in real time based on the view rotation sensitivity parameter. A quaternion interpolation algorithm is used to smoothly render the view's perspective transitions. For example, when a user quickly rotates the view, the quaternion interpolation algorithm makes the rotation transition more natural and smooth, avoiding stuttering or jumps. Furthermore, in actual implementation, the X / Y axis ratio of the original node size parameters is kept constant to ensure that the shape and proportion of the nodes remain unchanged during scaling and rotation, thus guaranteeing that users can accurately understand the connections and layout relationships between devices.

[0140] In a non-limiting embodiment, after outputting the real-time updated power grid operation and maintenance visualization interface, the method further includes: real-time acquisition of the rendering frame rate of the interactive visualization element set; activating a rendering load balancing mode when the frame rate is lower than a preset threshold; in the rendering load balancing mode, statistically analyzing the screen space ratio of the size parameters of each node in the dynamic topology graph, and filtering micro-nodes with a ratio less than a preset percentage threshold; replacing the node color wheel gradient parameters of the micro-nodes with sub-resolution texture maps, and merging the geometric meshes of adjacent micro-nodes into a target instantiation model; recalculating the collision volume of the bounding boxes of color blocks in the heatmap based on the target instantiation model to reduce the spatial segmentation granularity of the bounding boxes; using a culling algorithm to remove invisible color blocks outside the collision volume, generating an optimized visualization element set and triggering interface redraw. In actual implementation, micro-nodes that are focused by the user's touch can be excluded.

[0141] After outputting the real-time updated power grid operation and maintenance visualization interface, a rendering load balancing mechanism is introduced to ensure the interface's smoothness and efficiency. The system collects the rendering frame rate of the interactive visualization element set in real time, for example, once per second. For example, the preset threshold is 60 frames per second. When the collected frame rate is lower than this preset threshold, it means that the system's rendering load is heavy, and the rendering load balancing mode is activated at this time.

[0142] In rendering load balancing mode, the system first calculates the screen space percentage of each node's size parameters in the dynamic topology graph. For example, it determines the screen space percentage by calculating the pixel area occupied by each node on the screen. Then, it filters out micro-nodes whose percentage is less than a preset threshold; these micro-nodes are generally less visually important to the overall information display.

[0143] For these micro-nodes, the system replaces their node color wheel gradient parameters with sub-resolution texture maps. While sub-resolution texture maps are less detailed than the original high-resolution maps, they significantly reduce the resources required for rendering. Simultaneously, the system merges the geometric meshes of adjacent micro-nodes into a target instantiated model. For example, merging several adjacent and functionally similar micro-nodes into a larger model reduces the number of objects that need to be processed during rendering, thus improving rendering efficiency.

[0144] Next, the collision volume of the bounding boxes of color blocks in the heatmap is recalculated based on the target instantiated model. Due to the merging of micronodes and parameter adjustments, the bounding boxes of color blocks in the heatmap will also change. Recalculating the collision volume allows for a more accurate determination of which color blocks are visible and which are not from the current viewpoint. Simultaneously, reducing the spatial segmentation granularity of the bounding boxes means reducing the computational load for collision detection, further improving rendering efficiency.

[0145] Finally, the system employs a culling algorithm to remove invisible color blocks outside the collision volume. For example, by determining whether a color block is within the user's current visual range, those outside the visual range are directly removed from the rendering queue. This significantly reduces the resources required for rendering the resulting optimized set of visual elements, triggering a system-triggered interface redraw. In practical implementation, to avoid impacting the user experience, micro-nodes that are focused on by the user are excluded. These focused micro-nodes are typically the focus of user attention; preserving their original display ensures users can access detailed information. Simultaneously, by optimizing other non-focused micro-nodes, the overall interface rendering smoothness is guaranteed, allowing users to perform power grid operation and maintenance monitoring and analysis in an efficient and smooth visualization environment.

[0146] It is worth mentioning that, in the specific implementation process, those skilled in the art can perform the following processing based on existing technology to ensure the complete and clear implementation of the embodiments: improving the unit confusion problem by explicitly defining the temperature gradient change rate as a composite dimension ℃ / (m·min); improving the lack of basis for voltage safety threshold setting by referencing the permissible voltage deviation range (±10%) for low-voltage distribution systems specified in the International Electrotechnical Commission standard IEC60038; improving the time alignment accuracy problem of multi-source data by using a cubic spline interpolation algorithm to compensate for millisecond-level timestamp deviations; improving the insufficient model credibility problem by introducing cross-validation and grid search parameter tuning to optimize the environmental weight prediction model; improving the risk of feature distortion caused by a fixed normalization interval by using a dynamic sliding window extreme value method to adjust the normalization range in real time; and improving the accuracy by defining a gesture type priority mapping table (e.g., pinch to zoom > slide to pan > long press to focus). The following methods are proposed to improve the disordered response in user operation conflict scenarios: Improving the long-term bias accumulation problem of prediction models by establishing a regular online calibration mechanism between predicted equipment health values ​​and actual maintenance records; Improving the applicability error of similar equipment matrices when historical data is missing by constructing a 3D similarity matching algorithm based on equipment model, installation environment, and operating years; Improving the subjectivity problem of meteorological event alarm threshold setting by balancing false alarm and false negative rates through ROC curve analysis; Improving the rendering performance bottleneck problem in high-density node scenes by combining OpenGL multi-sampling anti-aliasing (MSAA) technology with instantiated rendering; Improving the misoperation recognition problem caused by touch parameter fluctuations by extracting gesture trajectory curvature features through a Bézier curve fitting algorithm; and Improving the stereoscopic perspective distortion problem caused by linear calculation of depth offset by establishing a nonlinear piecewise function mapping relationship between the Z-axis scaling factor and the average contact pressure.

[0147] In summary, this invention, through real-time acquisition of multi-source heterogeneous operation and maintenance data sets from smart grids, can comprehensively obtain various relevant information; cross-modal cleaning and time alignment processing can effectively improve data quality and consistency, making the preprocessed data set more usable; dynamic feature fusion and correlation analysis can deeply mine potential relationships in the data, thereby generating a feature set that accurately reflects equipment status and equipment-environment interaction; furthermore, based on a multi-dimensional visualization strategy, an interactive visualization element set can be generated, providing users with diverse interaction methods; finally, dynamic rendering and layout optimization can ensure real-time updates and a reasonable layout of the visualization interface. In addition, configuring a user operation priority response mechanism can ensure orderly responses to user operations in complex interaction scenarios. Thus, the efficiency and intelligence of power grid operation and maintenance visualization can be improved, thereby enhancing the timeliness of power grid operation and maintenance decision-making and management.

[0148] Based on the same inventive concept, embodiments of the present invention also provide an operation and maintenance data visualization system. (See also...) Figure 2 As shown, it is a schematic diagram of the structure of a possible operation and maintenance data visualization system provided in an embodiment of the present invention. Figure 2 In the system, the operation and maintenance data visualization system 200 includes a processor 210 and a memory 220. The memory 220 stores computer programs that can be executed by the processor 210. By executing the instructions stored in the memory 220, the processor 210 can perform the steps of the above-mentioned operation and maintenance data visualization method based on smart grid.

[0149] Based on the same inventive concept, embodiments of the present invention provide a computer-readable storage medium including a computer program. When the computer program is run on an operation and maintenance data visualization system, the computer program is used to cause the operation and maintenance data visualization system to perform the steps of the aforementioned smart grid-based operation and maintenance data visualization method. In some possible embodiments, various aspects of the smart grid-based operation and maintenance data visualization method provided by the present invention can also be implemented in the form of a program product, including a computer program. When the program product is run on an operation and maintenance data visualization system, the computer program is used to cause the operation and maintenance data visualization system to perform the steps in the aforementioned smart grid-based operation and maintenance data visualization method. For example, the operation and maintenance data visualization system can perform actions such as... Figure 1 The steps are shown in the figure.

[0150] In the technical solutions involved in the above embodiments of the present invention, whether performing comparison calculations of multi-dimensional features or constructing composite parameters, if there are problems caused by significant differences in the number of dimensions, units of measurement, and semantic meanings of different features, those skilled in the art, based on their professional knowledge and past practical experience, can fully understand that these differences need to be properly handled so that the calculation results are accurate and comparable, and to avoid situations such as logical confusion and unclear mathematical meaning.

[0151] In detail, when faced with features of different numbers of dimensions, in order to accurately calculate the similarity, matching degree or feature distance between different features, those skilled in the art can use a variety of strategies, such as feature selection, feature extraction, kernel function and other strategies for adaptive processing.

[0152] In order to achieve comparability alignment of feature spaces when processing the comparison of multidimensional features, those skilled in the art can use a variety of existing common technical means, including but not limited to the following existing technologies: standardization preprocessing, mapping transformation, spatial projection, etc.

[0153] In the process of constructing composite parameters (such as loss function values), different parameter terms often have different dimensions. Those skilled in the art can use existing normalization processing or adaptive weight allocation mechanisms based on distribution characteristics.

[0154] The aforementioned general techniques for solving feature matching and loss balance problems are all common knowledge in this field. These techniques have been fully verified and widely used in numerous practical applications, and those skilled in the art can skillfully and flexibly apply these methods to handle similar problems involving differences in dimensions.

[0155] The formulas and calculation processes involved in the embodiments of this invention, whether used for multidimensional feature comparison or composite loss function construction, strictly adhere to the principle of dimensional correspondence. Each variable in the formula has a clear and explicit physical meaning, and its operational logic fully conforms to basic mathematical and physical logic. The calculation results are necessarily the reasonable results expected by this invention. Those skilled in the art are capable of effectively solving various problems arising from the number of dimensions, dimensional differences, etc., in the multidimensional feature comparison calculation and composite loss function construction in the embodiments, based on specific data conditions and business needs, by comprehensively utilizing the above-mentioned general technical means, thus ensuring the accuracy, reliability, and implementability of the technical solution of this invention.

Claims

1. A smart grid-based operation and maintenance data visualization method, characterized in that, The method comprises the following steps: real-time collection of multi-source heterogeneous operation and maintenance data sets in a smart power grid; cross-modal cleaning and time alignment processing of the operation and maintenance data set to generate a preprocessed data set; dynamic feature fusion and correlation analysis of the preprocessed data set to generate a device state feature set and a device-environment interaction feature set; generation of an interactive visualization element set through multi-dimensional visualization strategies based on the device state feature set and the device-environment interaction feature set; dynamic rendering and layout optimization of the interactive visualization element set to output a real-time updated power grid operation and maintenance visualization interface; wherein the power grid operation and maintenance visualization interface is configured with a user operation priority response mechanism for handling interaction conflicts; periodic reading of node color ring gradient parameters in the interactive visualization element set, and statistics of the occurrence frequency of each node color ring gradient parameter within a preset time window; processing of the occurrence frequency through a preset device state degradation model to output device health prediction values corresponding to different color ring gradients; mapping the device health prediction values to a filling ratio parameter of a ring progress bar, and generating a device health label according to the filling ratio parameter; dynamic binding of the device health label in the peripheral area of the dynamic topology graph, and updating the edge transition state of the node color ring gradient parameter through anti-aliasing rendering.

2. The method of claim 1, wherein, The cross-modal cleaning and time alignment processing of the operation and maintenance data set to generate a preprocessed data set comprises: dividing the current time series data of the operation and maintenance data set into initial data segments according to a preset time granularity, and generating interpolated current data segments by linear interpolation based on adjacent data trends for missing initial data segments; detecting abnormal voltage points in the voltage data segment of the operation and maintenance data set that exceed a preset safety threshold, and replacing the abnormal voltage points with the historical voltage mean value of the corresponding device to generate a corrected voltage data segment; performing sliding window mean filtering on the temperature gradient data of the operation and maintenance data set, identifying abnormal fluctuation intervals with a temperature change rate exceeding a threshold within the window, and generating denoised temperature data segments using Gaussian filtering; extracting the timestamp information of the meteorological correlation data of the operation and maintenance data set, synchronizing the timestamp information of the meteorological correlation data with the timestamps of the interpolated current data segments, the corrected voltage data segments, and the denoised temperature data segments to generate a synchronized environmental influence factor sequence; unifying the time coding of the interpolated current data segments, the corrected voltage data segments, the denoised temperature data segments, and the synchronized environmental influence factor sequence to generate the preprocessed data set.

3. The method of claim 2, wherein, The dynamic feature fusion and correlation analysis of the preprocessed data set to generate a device state feature set and a device-environment interaction feature set comprises: multi-scale wavelet decomposition of the interpolated current data segments to generate fluctuation components and trend components, and calculation of the energy density ratio of each component; density clustering of the corrected voltage data segments to generate abnormal voltage point distribution density, and calculation of regional cumulative anomaly index according to the clustering results; Performing slope analysis on the noise-reduced temperature data segment to extract the temperature rising rate, and counting the number of times that the temperature fluctuation exceeds the threshold within a unit time to generate an abnormal fluctuation frequency; Inputting the synchronized environmental influence factor sequence into a pre-trained environmental weight prediction model to output a combination vector of temperature influence weight, humidity influence weight, and wind speed influence weight; Concatenating the energy density ratio, the abnormal voltage point distribution density, the regional cumulative anomaly index, the temperature rising rate, and the combination vector to generate a device state feature set and a device-environment interaction feature set.

4. The method of claim 3, wherein, Based on the device state feature set and the device-environment interaction feature set, an interactive visualization element set is generated through a multi-dimensional visualization strategy, including: Using a multi-view projection strategy to map the energy density ratio and the regional cumulative anomaly index to a three-dimensional visualization space, wherein the node size parameter of the dynamic topology graph is generated based on the energy density ratio, and the node color ring gradient parameter is generated based on the regional cumulative anomaly index; Based on a hierarchical mapping strategy, the temperature rising rate is mapped to generate a three-dimensional heat map color block altitude parameter through a height field, and a color block pulsation parameter is generated through a transparency decay function based on the abnormal fluctuation frequency; Using a multi-dimensional vector fusion algorithm, the temperature influence weight, humidity influence weight, and wind speed influence weight in the combination vector are respectively mapped to the arrow length parameter, feathering parameter, and inclination parameter of the meteorological vector; A cross-view association mechanism is established to synchronize and align the node size parameter and the node color ring gradient parameter of the dynamic topology graph, the color block altitude parameter and the color block pulsation parameter of the three-dimensional heat map, and the arrow length parameter, feathering parameter, and inclination parameter of the meteorological vector in the spatial coordinate, generating an interactive visualization element set with coordinate linkage.

5. The method of claim 4, wherein, The interactive visualization element set is dynamically rendered and layout optimized to output a real-time updated power grid operation and maintenance visualization interface, including: Based on the node size parameter and the node color ring gradient parameter of the dynamic topology graph in the interactive visualization element set, a topology view in a three-dimensional space projection coordinate system is created; According to the color block altitude parameter and the color block pulsation parameter of the three-dimensional heat map, a heat view with a height field gradient is generated in the three-dimensional space projection coordinate system; The arrow length parameter, feathering parameter, and inclination parameter of the meteorological vector are mapped to the three-dimensional space projection coordinate system to generate a dynamic vector field meteorological view; According to the user terminal window size, an adaptive layout matrix is constructed, and the topology view, heat view, and meteorological view are spatially arranged according to a pre-set segmentation ratio to establish a coordinate mapping relationship between the three views; When a focus operation on a node region in the topology view is detected, the color block altitude parameter of the corresponding spatial coordinate in the heat view is synchronously adjusted based on the coordinate mapping relationship, and the feathering parameter transparency of the associated region vector in the meteorological view is improved; The motion trajectory of the meteorological vector is calculated by spatio-temporal interpolation, the trajectory smoothing parameter is generated based on the real-time change rate of the wind speed influence weight, and the dynamic rendering is performed using the Bezier curve fitting algorithm. An abnormal fluctuation propagation path is constructed in the thermal view, a path confidence interval is calculated according to the temperature rise rate and a historical propagation mode, and a gradient color light band is generated through a transparency gradient mapping; Real-time detection is performed on geometric collision relationships of elements in the dynamic topology graph, the three-dimensional thermal map and the meteorological view, and when a proportion of an intersection volume of a node bounding box of the dynamic topology graph and a color block bounding box of the three-dimensional thermal map to a smaller bounding box volume exceeds a preset percentage threshold, a spherical unfolding algorithm is activated to decouple an azimuth angle and recombine space of the collision view.

6. The method of claim 5, wherein, The method further comprises: A touch operation trajectory of a user on the power grid operation and maintenance visualization interface is monitored, and acceleration features and contact area change rates of the touch operation trajectory are extracted; A preset view scaling instruction set is matched according to the acceleration features, and a thermal zone focusing weight parameter is generated according to the contact area change rates; A scaling ratio of a node size parameter of the dynamic topology graph is adjusted based on the thermal zone focusing weight parameter, and a flicker frequency of an associated color block pulsation parameter in the three-dimensional thermal map is synchronously improved; The adjusted node size parameter and the color block pulsation parameter are input into a particle rendering module to generate a touch feedback particle flow, and the touch feedback particle flow is superimposed on an edge region of the power grid operation and maintenance visualization interface.

7. The method of claim 5, wherein, The method further comprises: A regional cumulative anomaly index in the device state feature set is analyzed in real time, and when the index exceeds a preset risk threshold, an anomaly tracing mode is activated; An association matrix of abnormal fluctuation frequencies and temperature rise rates of the same device in a historical operation and maintenance data set is extracted in the anomaly tracing mode; A probability density distribution graph of a device anomaly propagation path is generated according to the association matrix, and the probability density distribution graph is encoded as a semi-transparent thermal superimposed layer; The semi-transparent thermal superimposed layer is embedded into a lower region of a color block elevation parameter of the three-dimensional thermal map through a depth buffer layer rendering strategy.

8. The method of claim 5, wherein, The method further comprises: Time sequence change data of arrow length parameters and inclination parameters of the meteorological vector are continuously captured, and a dynamic attenuation curve of a wind speed influence weight is generated; Extreme point detection is performed on the dynamic attenuation curve, abnormal time segments with a slope mutation in the curve are extracted and marked as meteorological event trigger points; A temperature influence weight at a corresponding time in the synchronized environmental influence factor sequence is reversely searched according to the meteorological event trigger points; The temperature influence weight and the wind speed influence weight are normalized and fused, a meteorological event alarm identifier is generated, and the meteorological event alarm identifier is inserted into an arrow tail end of the dynamic vector field meteorological view.

9. An operations data visualization system, comprising: The device comprises a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method in any one of claims 1-8. The device comprises a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method in any one of claims 1-8.

Images