Substation equipment management platform combined with multi-source data fusion modeling

Abstract

The present invention discloses a substation equipment management platform combined with multi-source data fusion modeling, which relates to the field of substation management technology. The platform includes: an operation data set acquisition module for acquiring multi-source real-time operation data sets; an operation feature determination module for determining multiple operation features; a twin model construction module for constructing a three-dimensional digital twin model; an operation status information acquisition module for obtaining operation status information; an operation and maintenance analysis module for performing operation and maintenance analysis and formulating equipment management suggestions; and a management control module for intelligent management and control of equipment management suggestions. The present invention solves the technical problems of single data dimension, lagging status assessment and static operation and maintenance strategy in traditional substation equipment management in the prior art, and achieves the technical effect of realizing accurate assessment of substation equipment operation status and significantly improving equipment management efficiency through dynamic feedback optimization through multi-source real-time data acquisition and digital twin modeling.

Description

Technical Field

[0001] The present invention relates to the technical field of substation management, and in particular to a substation equipment management platform combined with multi-source data fusion modeling. Background Art

[0002] In the power system, substations serve as critical hubs connecting power generation and consumption, and the stable operation of their equipment is crucial. However, current substation equipment management faces numerous challenges. Traditional data collection methods are limited, relying on a limited number of sensors to capture a few key parameters. These methods, focusing solely on electrical quantities like voltage and current, fail to fully capture the equipment's operating status. This one-sided data collection results in a limited understanding of equipment operating conditions, preventing the timely detection of many potential faults. Furthermore, existing technologies lack efficient data fusion and deep analysis capabilities for analyzing operational characteristics. Multi-source data from different sensor types is difficult to effectively integrate and mine, resulting in limited statistical analysis and an inability to accurately identify key operational characteristics. This results in delayed and inaccurate warnings of equipment failures, increasing the risk of sudden equipment failures. Furthermore, operations and maintenance strategies rely primarily on manual experience. This approach is not only time-consuming and labor-intensive, but also prone to decision-making errors due to individual experience differences. Faced with complex and changing equipment operating conditions, manual troubleshooting and resolution is inefficient, resulting in prolonged equipment downtime and severely impacting the stability and reliability of power supply.

[0003] The existing technology has technical problems in traditional substation equipment management, such as single data dimension, lagging status assessment and static operation and maintenance strategy. Summary of the Invention

[0004] This application provides a substation equipment management platform that combines multi-source data fusion modeling to solve the technical problems of single data dimension, delayed status assessment and static operation and maintenance strategy in traditional substation equipment management in the existing technology.

[0005] In view of the above problems, this application provides a substation equipment management platform that combines multi-source data fusion modeling.

[0006] This application provides a substation equipment management platform that combines multi-source data fusion modeling, and the platform includes:

[0007] An operation data set acquisition module is used to collect real-time data through multiple data sources to obtain multi-source real-time operation data sets of multiple substation equipment in the target area; an operation feature determination module is used to fuse the multi-source real-time operation data sets to generate a fused data set, perform feature analysis based on the fused data set, and determine multiple operation features; a twin model construction module is used to perform three-dimensional modeling of multiple substation equipment based on Gaussian modeling technology to construct three-dimensional digital twin models of multiple substation equipment; an operation status information acquisition module is used to synchronize the fused data set to the three-dimensional digital twin model to perform operation evaluation on multiple substation equipment and obtain equipment operation status information; an operation and maintenance analysis module is used to traverse the equipment operation status information and combine the multiple operation features to perform operation and maintenance analysis on multiple substation equipment and formulate equipment management recommendations; a management and control module is used to perform simulated operation and maintenance based on the equipment management recommendations, generate simulated operation and maintenance results, perform dynamic feedback optimization on the equipment management recommendations according to the simulated operation and maintenance results, generate an equipment management plan, and execute the equipment management plan to perform intelligent management and control of substation equipment.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] The system includes an operation dataset acquisition module, which collects real-time data from multiple data sources to obtain multi-source real-time operation datasets. The operation feature determination module fuses the multi-source real-time operation datasets to generate a fused dataset, performs feature analysis, and determines multiple operation features. The twin model construction module performs three-dimensional modeling on multiple substation devices to construct three-dimensional digital twin models of the devices. The operation status information acquisition module evaluates the operation of multiple substation devices and obtains device operation status information. The operation and maintenance analysis module traverses the device operation status information and combines the multiple operation features to perform operation and maintenance analysis on the multiple substation devices and formulate equipment management recommendations. The management and control module simulates operation and maintenance based on the equipment management recommendations, generates simulation operation and maintenance results, performs dynamic feedback optimization based on the simulation operation and maintenance results, generates equipment management plans, and performs intelligent management and control. This system achieves the technical effect of accurately evaluating the operation status of substation equipment and significantly improving equipment management efficiency through dynamic feedback optimization through multi-source real-time data acquisition and digital twin modeling. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0011] Figure 1 A schematic diagram of the structure of a substation equipment management platform combined with multi-source data fusion modeling provided in an embodiment of the present application;

[0012] Figure 2 A flow chart of the operation status information acquisition module in the substation equipment management platform combined with multi-source data fusion modeling provided in an embodiment of the present application.

[0013] Explanation of the accompanying drawings: operation data set acquisition module 10, operation feature determination module 20, twin model construction module 30, operation status information acquisition module 40, operation and maintenance analysis module 50, management control module 60. DETAILED DESCRIPTION

[0014] This application provides a substation equipment management platform that combines multi-source data fusion modeling to solve the technical problems of single data dimension, lagging status assessment and static operation and maintenance strategy in traditional substation equipment management in the existing technology.

[0015] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making any creative work are within the scope of protection of this application.

[0016] Examples, such as Figure 1 As shown, the present application provides a substation equipment management platform combined with multi-source data fusion modeling, the platform including:

[0017] The operation data set acquisition module 10 is used to collect real-time data from multiple data sources to obtain multi-source real-time operation data sets of multiple substation equipment in a target area.

[0018] Specifically, the operating data set acquisition module 10 collects real-time operating data sets from multiple sources across multiple substations within the target area. On the hardware level, it is equipped with high-performance data acquisition cards and multi-port network adapters to ensure simultaneous connection to multiple data sources. For sensor data sources, industrial-grade wired or wireless sensor network technologies, such as ZigBee and LoRa, are used to establish reliable communication links with various sensors to acquire analog and digital data such as temperature, pressure, and vibration. A / D conversion modules are used to convert analog signals into digital signals. On the software side, a dedicated data acquisition program has been developed that supports multiple communication protocols, such as Modbus and OPC UA, for data exchange with automated monitoring systems and other intelligent devices. Data parsing rules and caching mechanisms are tailored to the characteristics of different data sources. For data sources with high real-time requirements, such as online monitoring sensors, real-time data push technology is used to ensure rapid data transmission to the module. For relatively static data sources, such as historical maintenance records, regular data pull and update operations ensure data timeliness. At the same time, in order to ensure the accuracy and integrity of the data, data verification and error correction functions are integrated, and the collected data is verified using the cyclic redundancy check (CRC) algorithm. Once erroneous data is found, it is immediately processed through data retransmission or error repair algorithm, thereby successfully obtaining high-quality multi-source real-time operation data sets.

[0019] The operation feature determination module 20 is configured to fuse the multi-source real-time operation data sets to generate a fused data set, perform feature analysis based on the fused data set, and determine a plurality of operation features.

[0020] Specifically, after receiving the multi-source real-time operational datasets from the operational dataset acquisition module, the operational feature determination module 20 first applies advanced time series analysis algorithms to accurately time-stamp each data point in the dataset, thereby determining the operational timestamp and establishing a precise time coordinate axis for the data, ensuring the orderliness of subsequent processing. Based on these timestamps, an alignment algorithm is used to sequentially arrange the multi-source data to generate operational series data columns. On this basis, the data columns are reconstructed using autoencoder technology from deep learning. By learning the inherent structure and characteristics of the data, the autoencoder encodes and fuses the multi-source real-time operational datasets into a more representative fused dataset. A comprehensive feature analysis is then performed on the fused dataset. Multivariate statistical analysis methods such as principal component analysis (PCA) and singular value decomposition (SVD) are used to reduce the data dimension and extract features, uncovering key information hidden within the data. At the same time, combined with spectrum analysis technology, data is analyzed from a frequency domain perspective to obtain operating characteristics such as frequency components and harmonic characteristics. Association rule mining algorithms are also used to analyze the correlation between different variables, further determining multiple operating characteristics that can reflect the operating status of substation equipment, providing a solid basis for subsequent equipment evaluation and operation and maintenance decisions.

[0021] The twin model construction module 30 is used to perform three-dimensional modeling on multiple substation equipment based on Gaussian modeling technology to construct three-dimensional digital twin models of multiple substation equipment.

[0022] Specifically, the twin model construction module 30 comprehensively analyzes various detailed information about substation equipment, including multi-dimensional data such as the equipment's physical structure, mechanical parameters, and electrical characteristics. Some of this data is derived from the equipment's design documentation, while others are historical data accumulated during long-term operation. Based on the principles of Gaussian modeling, the equipment's geometric shape and spatial position relationships are abstracted into a Gaussian distribution. For example, the complex shapes of key components such as the core and windings in a substation are decomposed into multiple substructures based on Gaussian functions. By accurately calculating the parameters of these substructures, such as the mean and covariance, their positions and shapes in three-dimensional space are determined, gradually constructing a three-dimensional model of the entire transformer. Other equipment, such as circuit breakers and instrument transformers, are similarly modeled based on their respective structural characteristics and operating principles. During the modeling process, the models are continuously optimized and calibrated based on actual operating data to ensure that they accurately reflect the state changes of the equipment under different operating conditions. Ultimately, 3D digital twin models of multiple substation equipment were successfully constructed. These models not only provide an intuitive interface for visual equipment management but also lay a solid foundation for subsequent operational evaluation and fault prediction.

[0023] The operation status information acquisition module 40 is used to synchronize the fused data set to the three-dimensional digital twin model to perform operation evaluation on multiple substation equipment and obtain equipment operation status information.

[0024] Specifically, the operating status information acquisition module 40 first synchronizes the fused dataset completely and accurately to the 3D digital twin model through a specially designed data interface and efficient data transmission protocol. During the synchronization process, data verification and error correction are performed to ensure data reliability. Once the data is successfully imported into the model, a comprehensive assessment is conducted on each substation device based on key operating indicators. These key indicators include electrical parameters such as voltage and current stability, power factor, and other indicators; mechanical performance indicators such as vibration amplitude and speed stability; and temperature-related indicators, including the temperature range of key equipment components. For electrical parameters, a professional electrical analysis algorithm is used to calculate the deviation rate between the actual value and the standard value, and a quantitative score is assigned based on the degree of deviation. For mechanical performance, spectral analysis of the data collected by the vibration sensor is performed to determine the health of the equipment's mechanical structure and assign a corresponding assessment score. For temperature indicators, the temperature is analyzed to determine whether it is within the normal range based on the device's heat dissipation model and material thermal properties, thereby determining the score for this indicator. Combining these diverse assessment results ultimately generates detailed and accurate operating status information for each substation device, clearly reflecting the equipment's operational performance and providing a key basis for subsequent operation and maintenance decisions.

[0025] The operation and maintenance analysis module 50 is used to traverse the equipment operation status information and combine the multiple operation characteristics to perform operation and maintenance analysis on multiple substation equipment and formulate equipment management suggestions.

[0026] Specifically, during the data integration and preliminary analysis phase, the operation and maintenance analysis module 50 first uses the principal component analysis (PCA) algorithm to reduce the dimensionality of equipment operating status information and operating characteristic data, extracting key information and removing redundant data, thereby improving the efficiency and accuracy of subsequent analysis. To explore inherent data associations, the Apriori algorithm is used to mine association rules. By setting appropriate support and confidence thresholds, strong association rules between equipment operating status parameters and operating characteristics are identified. For example, it was found that when the temperature of a certain equipment component remains above a certain threshold and the operating time reaches a certain period, there is a high correlation with the failure risk of that component. To group equipment, the K-Means clustering algorithm is used to divide multiple substation equipment into clusters based on similarities in operating status and feature space. For each cluster, a decision tree-based classification algorithm is used to construct a decision tree model based on historical data and expert experience. Management recommendations are generated based on the characteristics of the equipment in each cluster. For example, for equipment in a cluster with a higher failure risk, the decision tree model recommends prioritizing in-depth inspection and preventive maintenance, thereby providing scientific and precise guidance for substation equipment operation and maintenance management.

[0027] The management control module 60 is used to perform simulated operation and maintenance based on the equipment management suggestions, generate simulated operation and maintenance results, perform dynamic feedback optimization on the equipment management suggestions according to the simulated operation and maintenance results, generate an equipment management plan, and execute the equipment management plan to perform intelligent management and control of substation equipment.

[0028] Specifically, during the simulated operation and maintenance phase, the management and control module 60 uses a Monte Carlo simulation algorithm to perform numerous simulations of equipment operation under various random operating conditions. Based on the equipment management recommendations, various input parameters, such as maintenance intervals and spare parts replacement strategies, are set. Through multiple simulations, the probability distribution of equipment performance indicators is determined, generating simulated operation and maintenance results. When comparing and analyzing the simulated operation and maintenance results with the expected targets, the mean square error (MSE) algorithm is used to calculate the deviation between the two. The effectiveness of the equipment management recommendations is determined based on the magnitude of the deviation. If the deviation exceeds a set threshold, an optimization process based on a genetic algorithm is initiated. In the genetic algorithm, key parameters in the equipment management recommendations are encoded as chromosomes. Selection, crossover, and mutation operations are used to continuously search for optimal parameter combinations to reduce the MSE value. After multiple rounds of iterative optimization, the optimal equipment management solution is determined. Real-time control algorithms, such as the proportional-integral-derivative (PID) algorithm, are then used to precisely adjust the equipment operating parameters based on feedback from the equipment's operating status, enabling intelligent management and control of substation equipment and ensuring optimal equipment performance.

[0029] In one possible implementation, the operation characteristic determination module 20 further includes:

[0030] The operation timestamp determining unit is used to perform operation timing analysis on the multiple substation devices according to the multiple data sources to determine the operation timestamp.

[0031] The runtime data sequence generating unit is configured to align the multi-source real-time runtime data sets according to the runtime timestamps to generate a runtime data sequence.

[0032] An autoencoder construction unit is used to traverse the runtime series data column to reconstruct the multi-source real-time runtime data set and construct an autoencoder.

[0033] The encoding fusion unit is used to encode and fuse the multi-source real-time running data sets through the autoencoder to generate the fused data set.

[0034] Specifically, a time series analysis algorithm is used to analyze the massive amount of data about multiple substation equipment provided by multiple data sources. By identifying the time mark information in the data or based on the order in which the data is generated, the operating timestamp corresponding to each data point can be accurately determined, so that the data can be arranged in order in the time dimension.

[0035] Based on the determined runtime timestamps, the multi-source real-time runtime datasets are aligned to generate a runtime series data column. Using a specific data source as a benchmark, typically one with high data integrity and accuracy, the data source is selected. Using the timestamp sequence of the benchmark data source as a clue, matching timestamps are searched for in other data sources. For timestamps with exact matches, the corresponding data is sorted according to pre-set rules (such as data source importance and data type) and inserted into the corresponding position in the runtime series data column. If there are slight differences in timestamps within an acceptable error range, linear interpolation is used to fit the data. Each data point is verified by calculation and comparison with pre-set values. If the verification fails, indicating that the data may be erroneous or missing, it is handled according to a pre-defined strategy, such as replacing it with the nearest correct data or issuing an anomaly alert for manual processing. Ultimately, a qualified runtime series data column is generated.

[0036] The runtime data series is traversed, and for each time point, relevant information is extracted from multiple data sources to form a feature vector. For example, corresponding values ​​are extracted from data sources such as voltage, current, and power, forming a multi-dimensional feature vector. These feature vectors serve as the input data for the autoencoder. When building an autoencoder, the encoder is built first. The encoder consists of a multi-layer neural network. The number of neurons in the input layer matches the dimension of the feature vector and is responsible for receiving the input data. As the network layer deepens, the number of neurons is gradually reduced to achieve data compression and extract key features. For example, a fully connected layer is used to multiply the input data by a weight matrix, add a bias term, and then process it through an activation function (such as the ReLU function) to obtain a compressed feature representation. The decoder corresponds to the encoder and also consists of a multi-layer neural network. The number of neurons in its input layer matches the dimension of the feature representation output by the encoder, and the number of neurons in its output layer matches the dimension of the original input feature vector. The decoder performs the reverse operation to gradually restore the compressed features output by the encoder to a data form similar to the original input. Similarly, fully connected layers, weight matrices, bias terms, and activation functions (such as linear activation functions) are used to attempt to reconstruct the original feature vector. Throughout the entire process, the encoder and decoder parameters (weights and biases) are continuously adjusted to minimize the difference between the reconstructed data and the original input data. This difference is typically measured using a loss function such as mean squared error. Optimization algorithms such as gradient descent, based on feedback from the loss function, iteratively update the parameters, enabling the autoencoder to learn the inherent structure and characteristics of the data, thereby completing the reconstruction of multi-source real-time datasets and the construction of the autoencoder.

[0037] After the autoencoder is constructed, it uses the autoencoder to perform encoding fusion on the multi-source real-time dataset, generating a fused dataset. Each data sample in the multi-source real-time dataset is sequentially input into the encoder portion of the constructed autoencoder. Based on its trained weights and biases, the encoder performs a complex nonlinear transformation on the input data, compressing the high-dimensional, multi-source raw data into low-dimensional feature vectors. These feature vectors contain key information about the original data. These encoded feature vectors are then passed to the decoder portion within the autoencoder. However, the goal here is not to decode them back to their original form; instead, the feature vectors output by the encoder are used for fusion. Specifically, the feature vectors encoded from different data sources are merged according to a specific rule. For example, the feature values ​​of corresponding dimensions can be simply concatenated, or multiple feature vectors can be fused into a single unified feature vector through weighted summation. In this way, all data samples in the multi-source real-time dataset are encoded and fused, ultimately generating a series of fused feature vectors that constitute the fused dataset. This fused dataset not only integrates information from multiple sources of data, but also removes redundant parts of the data through the encoding process of the autoencoder, making the data more representative and compact, providing a high-quality data foundation for subsequent tasks such as equipment operation analysis and fault diagnosis based on this dataset.

[0038] In one possible implementation, the operation characteristic determination module 20 further includes:

[0039] The multidimensional data matrix acquisition unit is used to load the fused data set to perform multidimensional data integration and obtain a multidimensional data matrix.

[0040] The multi-domain feature signal acquisition unit is used to traverse the multidimensional data matrix and perform feature mining according to the runtime sequence data column to obtain multi-domain feature signals, wherein the multi-domain features include operation frequency domain feature signals and operation time domain feature signals.

[0041] The frequency domain component information determining unit is configured to decompose the operation frequency domain characteristic signal according to the operation frequency domain characteristic signal to determine a plurality of frequency domain component information.

[0042] The correlation analysis unit is used to traverse the plurality of frequency domain component information and perform correlation analysis with a plurality of substation equipment to obtain a plurality of equipment operation cycle characteristics.

[0043] The equipment operation cycle feature adding unit is configured to add the plurality of equipment operation cycle features to the plurality of operation features.

[0044] Specifically, when loading a fused dataset for multidimensional data integration to obtain a multidimensional data matrix, the fused dataset is first read from the storage device into memory using a file reading algorithm. Assume that the fused dataset is stored in a table format, with each row representing a data sample and each column representing a feature from a different data source. Next, an empty multidimensional array is initialized to store the integrated multidimensional data matrix. Each row of data in the fused dataset is then iterated over, and each eigenvalue in each row is processed based on the dimension category to which it belongs. For example, a dimension mapping table can be predefined that maps different feature names to corresponding dimension indices. The dimension mapping table determines the placement of each eigenvalue in the multidimensional array. If a feature belongs to the electrical parameter dimension, indexed as 0 in the multidimensional array, and another feature belongs to the environmental parameter dimension, indexed as 1, the electrical parameter eigenvalue is stored in column 0 of the corresponding row in the multidimensional array, and the environmental parameter eigenvalue is stored in column 1. When placing eigenvalues, data type consistency and compatibility must also be considered. Numerical data can be stored directly; non-numeric data, such as text labels, requires encoding conversion, such as using a label encoding algorithm to convert it to numeric form before storing it in a multidimensional array. After traversing all rows of data and performing operations such as data classification, placement, and type conversion, this multidimensional array forms the required multidimensional data matrix, providing a structured data foundation for subsequent data processing and analysis.

[0045] After obtaining the multidimensional data matrix, we begin traversing this matrix. Using the operational sequence data as a clue, we analyze each data point in the matrix sequentially. In the time domain, by observing how the data changes over time and using statistical methods such as calculating the mean, variance, peak value, and slope, we directly extract operational domain characteristic signals from the data sequence. These characteristic signals can reflect the fluctuations and stability of the equipment's operating status over time. In the frequency domain, we use algorithms such as Fourier transform to convert the time domain data into the frequency domain space. The transformed results are analyzed to extract information such as the amplitude and phase of different frequency components, thereby obtaining operational frequency domain characteristic signals. These signals reveal the energy distribution and characteristics of the equipment operating at different frequencies. Through a comprehensive traversal and analysis of the multidimensional data matrix, we ultimately successfully obtain multi-domain characteristic signals containing operational frequency domain characteristic signals and operational domain characteristic signals, providing a key basis for in-depth understanding of the operating status of substation equipment.

[0046] Based on the acquired operating frequency domain characteristic signal, it is used as the key basis for analyzing the operating frequency domain characteristic signal. Using short-time Fourier transform, it can segment the operating frequency domain characteristic signal with a certain window length in the time domain, and then perform Fourier transform on each signal segment, thereby converting the time domain signal into a two-dimensional time-frequency representation. In this two-dimensional representation, the amplitude and phase information corresponding to each frequency is the manifestation of the frequency component in the time domain signal. By extracting and analyzing the parameters such as the amplitude, phase and frequency value of these frequency components, several frequency domain component information is determined. This information describes in detail the composition structure of the operating frequency domain characteristic signal at different frequencies, providing detailed frequency dimension data support for further understanding of the equipment operation characteristics.

[0047] After obtaining several frequency domain component data, the system begins to traverse this information and perform correlation analysis with multiple substation devices. For each frequency domain component, the system first identifies the dominant frequency feature with the highest energy. Because the dominant frequency feature carries the signal's key information and reflects the core operating principles of the device, it serves as a key indicator for measuring the device's cyclical behavior. For each substation device, various operating data, such as fluctuations in voltage, current, and power, are collected over the same time span. A specialized correlation analysis algorithm is then used to calculate the Pearson correlation coefficient to quantify the correlation between the dominant frequency feature and each device's operating parameters. During the analysis, particular attention is paid to data variation patterns that show significant correlation with the dominant frequency feature. For example, if the current value of a device exhibits a stable, periodic correlation with the dominant frequency feature within a specific time interval, then the characteristics of this periodic variation, such as the time interval and amplitude variation, can be identified as part of the device's cyclical operating characteristics. By conducting such detailed analysis on each frequency domain component information and multiple substation equipment one by one, multiple equipment operation cycle characteristics are mined from a large amount of data association. These characteristics can effectively reveal the periodic laws in the equipment operation process and provide an important basis for equipment status monitoring, fault prediction, etc.

[0048] After successfully acquiring multiple equipment operation cycle characteristics, these characteristics are integrated into the existing multiple operation characteristics. Each equipment operation cycle characteristic is traversed, and for each characteristic, it is added to the appropriate operation characteristic set according to pre-set rules or data structures. For example, if the operation characteristics are stored in dictionary form, where different keys represent different types of operation characteristics, then the equipment operation cycle characteristics are added to the corresponding key-value pair list or data structure according to their attributes. In this way, all equipment operation cycle characteristics are integrated one by one into multiple operation characteristic systems, further enriching and improving the content of the operation characteristics, and providing more complete data support for the subsequent comprehensive analysis and evaluation of the operating status of substation equipment.

[0049] In one possible implementation, Figure 2 As shown, the operation status information acquisition module 40 also includes:

[0050] An operation evaluation index setting unit is used to synchronize the fused data set to the three-dimensional digital twin model, map the fused data set to multiple device operation components through the three-dimensional digital twin model, and set multiple operation evaluation indicators.

[0051] The operation score generating unit is used to perform operation evaluation on multiple substation equipment according to the multiple operation evaluation indicators to generate multiple operation scores.

[0052] The status label acquisition unit is used to divide the status into multiple levels, match and identify the multiple operation scores with the multiple status levels, and obtain multiple status labels.

[0053] The dynamic monitoring unit is used to dynamically monitor the multiple substation devices according to the multiple status tags and generate the device operation status information.

[0054] Specifically, the fused dataset is synchronized with a 3D digital twin model. This 3D digital twin model serves as a virtual representation of the actual substation equipment, accurately simulating its appearance, structure, and operational logic. After data synchronization, the fused dataset is mapped to multiple operational components within the model, ensuring that each component has access to relevant operational data. Furthermore, multiple operational evaluation indicators are established to measure equipment health across key dimensions such as performance, stability, and energy consumption, providing comprehensive and accurate benchmarks for subsequent evaluations.

[0055] For multiple substation equipment, operational evaluations are conducted based on multiple operational evaluation indicators, generating operational scores. First, for each equipment, relevant data for each indicator is extracted from the fused dataset. For example, for equipment performance indicators, output power and efficiency data are obtained. The output power score is calculated as follows: if the actual output power is lower than the ideal value, the ratio of the actual value to the ideal value is multiplied by a certain score. If it is higher than the ideal value, the excess is added to the base score and a proportional score is calculated. The efficiency score is calculated by multiplying the ratio of the actual value to the ideal value by a certain score. The two scores are added together to form the equipment performance score. For stability indicators, data on the frequency and amplitude of operating parameter fluctuations are extracted. The frequency score is calculated by multiplying the difference between the actual value and the ideal value by the corresponding score. The same applies to the amplitude. The two scores are added together to form the stability score. For energy consumption indicators, energy consumption per unit time is obtained. If the actual energy consumption is lower than the ideal value, the difference is multiplied by the full score. If it is higher than the ideal value, the energy consumption score is calculated by multiplying the difference between the actual value and the maximum acceptable value by the full score. Finally, based on the pre-defined weights for each indicator, the scores for equipment performance, stability, and energy consumption are weighted and summed to produce a comprehensive operational score for each equipment.

[0056] Based on industry standards, historical data, and operational experience, equipment status levels are categorized into five levels: excellent, good, qualified, warning, and fault. Each level is assigned a corresponding operational score range. For example, 90-100 points is considered excellent, 75-89 points is considered good, 60-74 points is considered qualified, 40-59 points is considered warning, and 40 points or less is considered fault. The operational score for each substation equipment is then matched and identified with the pre-defined status levels. For example, if a device's operational score is 85 points, the corresponding status label for that device is good. In this way, corresponding status labels are assigned to multiple substation equipment.

[0057] Based on these status tags, dynamic monitoring is conducted on multiple substation devices, continuously tracking changes in device status tags. Once a status tag changes, this information is immediately captured. For example, if a device's status tag was originally good, but an anomaly occurs during operation, causing the score to drop, the status tag changes to warning. This change is quickly identified, and detailed device operating status information is generated by combining basic device information, real-time operating data, and historical status records. This information not only includes the current device status, but also covers the time of the status change and possible causes of the change. This provides strong support for operation and maintenance personnel to promptly understand the equipment's operating status and take targeted measures.

[0058] In one possible implementation, the operation and maintenance analysis module 50 further includes:

[0059] The operation fault cycle status data generating unit is used to perform operation fault analysis on the equipment operation status information according to the multiple equipment operation cycle characteristics to generate operation fault cycle status data.

[0060] The cycle state cluster generating unit is used to perform cluster analysis based on the operation fault cycle state data, determine multiple cycle state cluster centers, divide the operation fault cycle state data according to the multiple cycle state cluster centers, and generate multiple cycle state clusters.

[0061] A multi-level equipment management unit is used to traverse the multiple periodic state clusters to perform management analysis on multiple substation equipment, determine multiple management modes, perform reinforcement learning based on the multiple management modes, obtain management learning results, perform multi-level equipment management based on the management learning results, and formulate the equipment management recommendations.

[0062] Specifically, after obtaining equipment operating status information, operational fault analysis is conducted based on multiple equipment operating cycle characteristics. By combining equipment operating status with operating cycle characteristics, the timing and frequency of fault occurrences, as well as the correlation between fault status and cycle characteristics, are analyzed to generate operational fault cycle status data. This data details the specific circumstances of equipment faults during their operating cycle, such as the distribution of faults within the cycle and the relationship between fault duration and cycle.

[0063] Cluster analysis of operational fault cycle state data was performed using the K-Means algorithm. By calculating the distances between data points, multiple cycle state cluster centers were determined. These cluster centers represent different types of fault cycle state patterns. The operational fault cycle state data was then divided according to these cluster centers, with similar fault cycle state data grouped into the same category. This generated multiple cycle state clusters, each of which reflects a specific equipment fault cycle pattern.

[0064] The system begins by traversing multiple cyclical state clusters and conducting a management analysis of multiple substation devices based on the failure cycle characteristics of the devices represented by each cluster. During the analysis, factors such as the device's characteristics, criticality, and operating environment determine whether to adopt a single management model (developing a dedicated management strategy for each device) or a cluster management model (grouping multiple devices with similar failure cycle characteristics and developing a unified management plan). For example, for critical and unique devices with significant failure impacts and unique operating characteristics, a single management model is adopted, with detailed, tailored inspection and maintenance plans. For a group of devices with the same model, similar operating environments, and similar failure cycle characteristics, a cluster management model is adopted to improve management efficiency. After determining multiple management models, reinforcement learning is used as a foundation. During the reinforcement learning process, various management scenarios are simulated, different management operations are repeatedly tested, and feedback is obtained based on a pre-defined reward mechanism. This reward mechanism measures the degree to which a management action improves the device's operating status. For example, if a management action reduces the device's failure rate or improves operational stability, a higher reward is given; otherwise, a lower reward is given. Through continuous trial and error and optimization, we gradually learn which management model achieves the best results under certain circumstances, thereby obtaining management learning results. Based on these management learning results, we further implement multi-level equipment management. At the individual equipment level, we consider the unique needs and operating conditions of each device and formulate personalized management recommendations, such as maintenance time and repair methods for specific equipment. At the equipment group level, we rationally allocate resources and arrange unified maintenance activities based on the overall operating characteristics and failure patterns of the equipment group. At the level of the entire substation system, we comprehensively consider the mutual influence between each equipment group and the overall operating objectives of the system, and conduct macro-level resource allocation and management coordination. Through this multi-level equipment management, we ultimately formulate comprehensive, detailed, and targeted equipment management recommendations to ensure the efficient and stable operation of substation equipment.

[0065] In one possible implementation, the multi-level device management unit further includes:

[0066] The change trend analysis unit is used to perform change trend analysis on a plurality of substation devices based on the plurality of periodic state clusters to obtain a plurality of periodic change trend data.

[0067] The single device information determination unit is used to associate and map the multiple periodic change trend data, extract the periodic change trend data with the same trend change and combine them to determine multiple device groups, extract the periodic change trend data with different trend change and identify them, and determine multiple single device information.

[0068] The single device management mode generating unit is used to generate a single device management mode according to the plurality of single device information in combination with corresponding periodic change trend data.

[0069] The cluster device management mode generating unit is configured to generate a cluster device management mode according to the plurality of device groups in combination with corresponding periodic change trend data.

[0070] The multiple management mode acquisition units are configured to integrate the single device management mode with the cluster device management mode to obtain the multiple management modes.

[0071] Specifically, for multiple cycle state clusters, a comprehensive change trend analysis is carried out on multiple substation equipment to obtain multiple cycle change trend data. Specifically, starting from each cycle state cluster, the operating data of the corresponding equipment in the cluster are arranged in time series, and the long-term trend, seasonal fluctuation and random fluctuation components in the data are identified with the moving average method in the time series analysis method. For the key operating parameters of the equipment, such as voltage, current, temperature, etc., their numerical changes in different cycle stages are observed respectively. For example, how the voltage gradually rises or falls in a complete cycle, and the amplitude of the fluctuation is analyzed. By continuously tracking and analyzing these parameters over multiple cycles, the laws and patterns of parameter changes are captured, and then multiple cycle change trend data that details the evolution of the equipment operating status over the cycle are generated. These data provide a key basis for subsequent in-depth understanding of the equipment operating status and prediction of the future status of the equipment.

[0072] After obtaining multiple periodic change trend data, these data are correlated and mapped. This means establishing a corresponding relationship between a large number of data to facilitate analysis and comparison. Each periodic change trend data is compared with other data one by one. For periodic change trend data with the same trend change, that is, those data that show consistent performance in key features such as change direction, amplitude, and period, they are selected and combined. Based on these combinations, devices with the same trend change are grouped together, thereby determining multiple device groups. At the same time, for those periodic change trend data with different trend changes, that is, data that are significantly different from other data in key features, they are specially identified. These identifiers can accurately correspond to the device that generated the data. Through these identifiers, multiple single device information is determined. This information covers the device number, type, and its unique periodic change trend characteristics.

[0073] Based on the identified individual device information and its corresponding periodic trend data, a dedicated management model can be developed for each individual device. For example, based on the unique operating trends of the device, specific inspection schedules and targeted maintenance measures can be arranged to form a single device management model.

[0074] For multiple device groups, we also combine their corresponding periodic trend data to generate a cluster device management model. Given that the devices within the group share similar trend changes, we can develop unified management strategies, such as centralized maintenance plans and resource allocation schemes, to improve management efficiency and reduce management costs.

[0075] Finally, the generated single device management model is integrated with the cluster device management model. Taking into account the characteristics of different devices and the efficiency and effectiveness of management, the advantages of these two management models are combined to ultimately obtain multiple comprehensive, flexible, and targeted management models, providing diverse and effective solutions for substation equipment management.

[0076] In one possible implementation, the multi-level device management unit further includes:

[0077] The management state space generating unit is configured to perform operation and maintenance management state learning on the single device management mode and the cluster device management mode according to a plurality of substation devices, so as to generate a management state space.

[0078] The management action space generating unit is configured to perform operation and maintenance management action learning on the single device management mode and the cluster device management mode according to a plurality of substation devices, so as to generate a management action space.

[0079] A management learning unit is used to associate and fuse the management state space with the management action space to generate a fusion solution space. When the fusion solution space is balanced, a reward function is introduced to perform management learning on the single device management mode and the cluster device management mode to generate the management learning result.

[0080] Specifically, each substation device is assigned a unique identifier for accurate tracking and identification. A data acquisition system collects various data during device operation in real time. For example, sensors installed on the equipment continuously collect operating parameter data such as voltage, current, and temperature, and record it at regular intervals (e.g., every second). Furthermore, a fault monitoring system records the time of equipment fault occurrence, fault code (corresponding to different fault types), and the time it took to repair the fault. For equipment maintenance history, a maintenance record database is used to capture the time, content (e.g., replacement parts, software upgrades), and maintenance personnel information for each maintenance session. For single-device management, the collected operating parameter data for each device is normalized to ensure that it falls within the same numerical range, facilitating subsequent analysis and comparison. For fault data, fault frequency is calculated as the number of times within a certain period (e.g., a month). Fault repair time is calculated as the time from fault occurrence to repair completion, and these data are also normalized. For maintenance history, maintenance intervals are measured as the number of days since the last maintenance, and this is also normalized. This normalized data is combined into a feature vector, representing the operational management status of the device at a specific moment in the single-device management mode. For clustered device management, data processing follows the same approach. However, due to cluster management, additional factors such as the device's relative position within the cluster and its connections with other devices need to be considered. For example, devices playing a key role in the cluster are given a higher weight. By analyzing inter-device communication data and collaborative operation data, the degree of inter-device connection is determined, and these factors are quantified and incorporated into the feature vector. The feature vectors generated for all substation devices in both single-device and clustered device management modes are aggregated, and a management state space is constructed using machine learning data structures, such as multidimensional arrays or specialized matrix storage formats. In this space, each vector represents the operational management status of a device under a specific management mode. The relationships between different vectors reflect the differences and similarities in the operational status of different devices, providing a comprehensive and structured data foundation for subsequent analysis and decision-making.

[0081] For multiple substation devices, we first comprehensively collect information on operation and maintenance management actions under both single-device and cluster device management modes through device operation log records and automated system command capture. These actions are then classified into categories such as routine inspections and fault repairs, and each specific action type is assigned a unique code. For the single-device management mode, the actions of each device are extracted based on the operation records, and action sequences are formed by time. These sequences are then converted into feature vectors using a label encoding algorithm. For the cluster device management mode, in addition to focusing on individual device actions, collaborative actions between devices are also analyzed, similarly forming action sequences and generating feature vectors. Finally, the feature vectors generated by multiple substation devices under both management modes are aggregated to construct a management action space in the form of a matrix or tensor, providing a structured data foundation for subsequent analysis.

[0082] The previously constructed management state space and management action space are linked and fused. During the fusion process, each specific action in the management action space is systematically matched to the corresponding various equipment states in the management state space. For example, for the action "equipment inspection," its results are analyzed when the equipment is in different states, such as "high-load operation," "normal operation," and "fault warning." Through a large number of such matching and analysis, a comprehensive understanding of the effects of each action in different states is obtained, which in turn generates a fused solution space that clearly demonstrates the various possible outcomes of the interaction between actions and states, providing a rich information foundation for subsequent management decisions.

[0083] After a series of analyses and integrations, when the fused solution space reaches equilibrium, this indicates that a relatively stable and coordinated relationship has been established between O&M actions and device status within the current O&M management system. To further optimize the management strategy, a reward function is introduced. This reward function is designed based on a set of pre-defined criteria and objectives. For example, if a management action significantly reduces device failure rates, improves operational stability, or enhances overall efficiency, then this action will receive a higher reward score according to the reward function. Conversely, if a management action causes more device problems or degrades performance, then it will receive a lower reward score. This reward mechanism enables continuous learning of various management actions in both single-device and cluster-based management modes. During this process, different management strategies are continuously tested and adjusted and optimized based on feedback from the reward function. For example, if a maintenance action targeting a specific device status in single-device management mode yields a high reward, the frequency of executing this action in similar situations is increased. Conversely, if a resource allocation strategy in cluster-based management mode results in a low reward score, adjustments to that strategy are attempted. After many such attempts and optimizations, management learning results are finally generated, which provide valuable reference for further improving the operation and maintenance management level of substation equipment, help formulate more scientific and efficient management plans, and ensure that substation equipment can operate stably and reliably.

[0084] In one possible implementation, the management control module 60 further includes:

[0085] The simulation scenario parameter construction unit is used to activate the simulation platform to model the operating environment of multiple substation equipment, construct simulation scenario parameters, map the equipment management suggestions to the simulation scenario parameters for simulation management, and generate multiple management effects.

[0086] The management effect judgment unit is used to set an expected management effect threshold according to the multiple status levels and judge whether the multiple management effects meet the expected management effect.

[0087] The operation and maintenance monitoring unit is used to generate a positive feedback signal if the multiple management effects meet the expected management effects, and continuously monitor the operation and maintenance of multiple substation equipment through the positive feedback signal to generate the simulated operation and maintenance results.

[0088] The abnormal operation and maintenance equipment set determination unit is used to generate a negative feedback signal if the multiple management effects do not meet the expected management effect, match and identify multiple substation equipment through the negative feedback signal, determine the abnormal operation and maintenance equipment set, and add the abnormal operation and maintenance equipment set to the simulation operation and maintenance result.

[0089] Specifically, after activating the simulation platform, it first collects physical environmental data (such as temperature, humidity, and air velocity) and electrical parameters (such as voltage, current, and load characteristics) within the substation using temperature and humidity sensors, transformers, and other devices. It also collects equipment technical parameters and aging information. Using CFD methods, signal processing techniques, and the development of equivalent circuits and aging models, it constructs simulation scenario parameters covering all aspects of equipment operation. It then analyzes the equipment management recommendations in detail, converting them into specific operational instructions and parameter adjustment requirements, and establishes a mapping table with the simulation scenario parameters. Parameters are adjusted within the simulation platform based on this table, such as simulating equipment maintenance or changing inspection cycles. The simulation is then started and time parameters are set. Multi-dimensional operational indicators such as equipment temperature, power loss, and failure rate are monitored in real time. Quantitative analysis is performed based on a predefined evaluation index system, ultimately generating multiple management performance data sets reflecting the effectiveness of implementing the management recommendations.

[0090] Based on multiple pre-defined status levels, and in conjunction with substation equipment's operational characteristics, industry standards, and O&M objectives, corresponding expected management effect thresholds are set for each status level. For example, for equipment in the "Excellent" status level, the expectation is for extremely high stability across key operational indicators, with specific thresholds set, such as a failure rate below 0.1% and an operating efficiency above 98%. For the "Good" status level, the failure rate thresholds might be set to below 0.5% and an operating efficiency above 95%. After obtaining multiple management effects, each indicator in each management effect is compared against the expected management effect threshold for the corresponding status level. For example, if the management effect for a particular piece of equipment involves indicators such as failure rate and operating efficiency, the actual failure rate and operating efficiency are compared against the failure rate and operating efficiency thresholds set for the equipment's status level. This determines whether the equipment's failure rate is below the set failure rate threshold and whether its operating efficiency is above the set efficiency threshold. By comprehensively comparing all management effect indicators with the expected management effect threshold, we can determine whether multiple management effects meet the expected management effects, thereby evaluating the effectiveness of equipment management suggestions in a simulation environment.

[0091] When it is determined that the multiple management effects meet the expected management effects, the mechanism for generating positive feedback signals is immediately triggered. This positive feedback signal acts as a positive instruction, quickly initiating the continuous operation and maintenance monitoring process for multiple substation equipment. In this process, with the help of monitoring technology and intelligent equipment, various operating parameters of substation equipment, such as voltage, current, power, temperature, frequency, etc., are collected and analyzed at high frequency and high precision. By comparing with historical data and preset standard parameters in real time, the changes in the equipment's operating status are analyzed to ensure that the equipment is always in a stable and efficient operating range. At the same time, these monitoring data are integrated and summarized to generate detailed and comprehensive simulation operation and maintenance results. These results not only cover the real-time operating status information of the equipment, but also include predictions of equipment performance trends, assessments of potential risks, and further verification of the effectiveness of existing management strategies. They provide a solid and reliable basis for subsequent operation and maintenance decisions, effectively ensuring the safe and stable operation of the substation.

[0092] When multiple management results deviate from the expected ones, a deviation analysis algorithm is used to generate a negative feedback signal. By calculating the difference between the actual management effect indicator and the expected management effect threshold, if the difference exceeds the preset tolerance range, the generation of a negative feedback signal is triggered. Next, a cluster analysis algorithm is used to process the operating data of multiple substation devices, grouping devices with similar operating characteristics and deviation trends. Based on this, a support vector machine (SVM) algorithm is used to classify and distinguish these clustered device data, accurately identifying those devices that deviate significantly from normal operating modes, thereby determining the set of abnormally operated and maintained devices. Finally, the information of these screened abnormal devices is integrated and added to the simulation operation and maintenance results, ensuring that detailed data for each abnormal device, such as device number, abnormal parameters, and the time of the abnormality, is fully recorded, providing a comprehensive and accurate basis for subsequent troubleshooting and optimization strategy formulation.

[0093] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0094] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

[0095] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.

Claims

1. The substation equipment management platform combined with multi-source data fusion modeling is characterized by: The platform includes: The operating data set acquisition module is used to collect real-time data from multiple data sources and obtain multi-source real-time operating data sets of multiple substation equipment in the target area; an operation feature determination module, configured to fuse the multi-source real-time operation data sets to generate a fused data set, perform feature analysis based on the fused data set, and determine a plurality of operation features; Twin model construction module, used to perform 3D modeling of multiple substation equipment based on Gaussian modeling technology and build 3D digital twin models of multiple substation equipment; An operation status information acquisition module, configured to synchronize the fused data set to the three-dimensional digital twin model to perform operation evaluation on multiple substation devices and obtain device operation status information; An operation and maintenance analysis module, configured to traverse the equipment operation status information and combine the multiple operation characteristics to perform operation and maintenance analysis on multiple substation equipment and formulate equipment management suggestions; A management and control module is configured to perform simulated operation and maintenance based on the equipment management suggestions, generate simulated operation and maintenance results, dynamically feedback and optimize the equipment management suggestions according to the simulated operation and maintenance results, generate an equipment management plan, and execute the equipment management plan to perform intelligent management and control of substation equipment; The operation characteristic determination module includes: an operation timestamp determining unit, configured to perform operation timing analysis on a plurality of substation devices according to the plurality of data sources to determine an operation timestamp; A runtime data sequence generating unit, configured to align the multi-source real-time runtime data sets according to the runtime timestamps to generate a runtime data sequence; An autoencoder construction unit is configured to traverse the runtime series data columns to reconstruct the multi-source real-time runtime data set and construct an autoencoder; an encoding fusion unit, configured to perform encoding fusion on the multi-source real-time running data sets through the autoencoder to generate the fused data set; a multidimensional data matrix acquisition unit, configured to load the fused data set to perform multidimensional data integration and obtain a multidimensional data matrix; a multi-domain feature signal acquisition unit, configured to traverse the multidimensional data matrix and perform feature mining according to the runtime sequence data column to obtain a multi-domain feature signal, wherein the multi-domain feature includes an operating frequency domain feature signal and an operating domain feature signal; a frequency domain component information determining unit, configured to decompose the operation frequency domain characteristic signal according to the operation frequency domain characteristic signal to determine a plurality of frequency domain component information; A correlation analysis unit, configured to perform correlation analysis on the plurality of frequency domain component information and the plurality of substation equipment to obtain operation cycle characteristics of the plurality of equipment; an equipment operation cycle feature adding unit, configured to add the plurality of equipment operation cycle features to the plurality of operation features; The operation and maintenance analysis module includes: an operation fault cycle state data generating unit, configured to perform operation fault analysis on the equipment operation state information according to the plurality of equipment operation cycle characteristics, and generate operation fault cycle state data; a periodic state cluster generating unit, configured to perform cluster analysis based on the operation fault periodic state data, determine a plurality of periodic state cluster centers, divide the operation fault periodic state data according to the plurality of periodic state cluster centers, and generate a plurality of periodic state clusters; A multi-level equipment management unit is used to traverse the multiple periodic state clusters to perform management analysis on multiple substation equipment, determine multiple management modes, perform reinforcement learning based on the multiple management modes, obtain management learning results, perform multi-level equipment management based on the management learning results, and formulate the equipment management recommendations.

2. The substation equipment management platform combined with multi-source data fusion modeling according to claim 1 is characterized in that: The operation status information acquisition module includes: an operation evaluation index setting unit, configured to synchronize the fused data set to the three-dimensional digital twin model, map the fused data set to a plurality of device operation components through the three-dimensional digital twin model, and set a plurality of operation evaluation indicators; An operation score generating unit, configured to perform operation evaluation on a plurality of substation devices according to the plurality of operation evaluation indicators and generate a plurality of operation scores; a status label acquisition unit, configured to divide the status into a plurality of levels, and perform matching identification on the basis of the plurality of operation scores and the plurality of status levels to obtain a plurality of status labels; The dynamic monitoring unit is used to dynamically monitor the multiple substation devices according to the multiple status tags and generate the device operation status information.

3. The substation equipment management platform combined with multi-source data fusion modeling according to claim 1 is characterized in that: The multi-level device management unit includes: a change trend analysis unit, configured to perform a change trend analysis on a plurality of substation devices based on the plurality of periodic state clusters to obtain a plurality of periodic change trend data; a single device information determination unit, configured to associate and map the plurality of periodic change trend data, extract periodic change trend data with the same trend change, combine them, determine a plurality of device groups, extract periodic change trend data with different trend change, identify them, and determine a plurality of single device information; a single device management mode generating unit, configured to generate a single device management mode based on the plurality of single device information in combination with corresponding periodic change trend data; A cluster device management mode generating unit, configured to generate a cluster device management mode according to the plurality of device groups in combination with corresponding periodic change trend data; The multiple management mode acquisition units are configured to integrate the single device management mode with the cluster device management mode to obtain the multiple management modes.

4. The substation equipment management platform combined with multi-source data fusion modeling according to claim 3 is characterized in that: The multi-level device management unit further includes: A management state space generating unit is configured to perform operation and maintenance management state learning on the single device management mode and the cluster device management mode according to a plurality of substation devices to generate a management state space; a management action space generating unit, configured to perform operation and maintenance management action learning on the single device management mode and the cluster device management mode according to a plurality of substation devices, and generate a management action space; A management learning unit is used to associate and fuse the management state space with the management action space to generate a fusion solution space. When the fusion solution space is balanced, a reward function is introduced to perform management learning on the single device management mode and the cluster device management mode to generate the management learning result.

5. The substation equipment management platform combined with multi-source data fusion modeling according to claim 2 is characterized in that: The management control module includes: A simulation scenario parameter construction unit is used to activate the simulation platform to model the operating environment of multiple substation devices, construct simulation scenario parameters, map the device management suggestions to the simulation scenario parameters to perform simulation management, and generate multiple management effects; a management effect judging unit, configured to set an expected management effect threshold according to the multiple status levels, and judge whether the multiple management effects meet the expected management effect; an operation and maintenance monitoring unit, configured to generate a positive feedback signal if the multiple management effects meet the expected management effects, and perform continuous operation and maintenance monitoring on the multiple substation equipment through the positive feedback signal to generate the simulated operation and maintenance result; The abnormal operation and maintenance equipment set determination unit is used to generate a negative feedback signal if the multiple management effects do not meet the expected management effect, match and identify multiple substation equipment through the negative feedback signal, determine the abnormal operation and maintenance equipment set, and add the abnormal operation and maintenance equipment set to the simulation operation and maintenance result.

Images