Monitoring and early warning method and system for power equipment of transformer substation

Abstract

The invention relates to the technical field of monitoring and early warning of power equipment, in particular to a monitoring and early warning method and system for substation power equipment. Comprising a data acquisition module, an intelligent diagnosis module, a risk assessment module, a collaborative early warning module and an interaction control module. The data acquisition module acquires equipment operation data including current, voltage, temperature and humidity and vibration signals. And the intelligent diagnosis module identifies equipment abnormity through the generative adversarial network and the graph neural network. The risk assessment module generates a risk assessment curve based on a fractional order differential model and a reinforcement learning algorithm. And the collaborative early warning module generates a graded early warning scheme by utilizing edge calculation. And the interaction control module displays a state and optimizes a control strategy through a human-computer interface. The intelligent monitoring and management level of the power equipment of the transformer substation is improved, the service life of the equipment is prolonged, and the operation and maintenance cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of monitoring and early warning of electric power equipment, and in particular to a monitoring and early warning method and system for electric power equipment in a substation. Background Art

[0002] With the rapid development of the power industry and the continuous expansion of the power grid, substations are important nodes in the power grid, and their safe operation directly affects the reliability and stability of the power system. The power equipment in the substation (such as transformers, circuit breakers, capacitors, etc.) is a key component of the power system operation. However, due to the complex operating environment, aging equipment or improper operation, substation power equipment often has hidden dangers of failure, such as overheating, partial discharge, mechanical damage, etc. If these hidden dangers are not discovered and handled in time, they may lead to equipment failure or even large-scale power outages in the power grid, causing significant economic losses and social impacts.

[0003] In the existing technology, traditional substation monitoring mainly relies on manual inspection and regular maintenance. However, manual inspection has defects such as long detection cycle, poor real-time performance, and insufficient coverage, making it difficult to timely discover potential problems in equipment operation. In addition, although some equipment is equipped with a single parameter monitoring function (such as temperature, humidity, current, etc.), it fails to achieve multi-parameter comprehensive analysis and fault warning, resulting in a low rate of hidden danger detection and difficulty in accurately locating the specific location of the fault; the evaluation model of the existing technology has low accuracy and lacks dynamic adjustment capabilities, making it difficult to reflect changes in equipment risks in real time. Summary of the invention

[0004] To solve the above problems, the present invention provides a monitoring and early warning method and system for substation power equipment, which solves the problem of how to achieve real-time monitoring and multi-parameter comprehensive analysis of the operating status of substation power equipment, timely identify potential fault hazards and accurately locate the fault site, and at the same time improve the dynamic adaptability of risk assessment to ensure the safe and reliable operation of the power system, thereby improving the intelligent monitoring and management level of substation power equipment.

[0005] To achieve the above object, the technical solution adopted by the present invention is: On the one hand, a monitoring and early warning system for power equipment in a substation includes a data acquisition module, an intelligent diagnosis module, a risk assessment module, a collaborative early warning module, and an interactive control module that are sequentially communicatively connected; The data acquisition module is used to collect the operating data of the power equipment in the substation, and the operating data includes the current, voltage, temperature, humidity and vibration signals of the power equipment; The intelligent diagnosis module is used to generate a diagnosis report for the power equipment by using a generative adversarial network combined with a graph neural network algorithm to perform in-depth analysis and abnormal pattern recognition on the operation characteristics of the power equipment based on the operation data; The risk assessment module is used to perform high-precision modeling of the rate of change and trend of the operating state of the power equipment based on the power equipment diagnostic report and a dynamic assessment framework combining a fractional-order differential model with a reinforcement learning algorithm, and dynamically optimize the equipment risk level assessment index using reinforcement learning to generate a real-time risk assessment curve; The collaborative warning module is used to perform multi-node collaborative processing on the real-time risk assessment curve by building a collaborative architecture based on edge computing, and generate a graded warning plan according to the area where the equipment is located and the operation level; The interactive control module is used for adaptive optimization algorithm based on fuzzy neural network, displays equipment operation status, fault information and early warning suggestions through human-computer interaction interface, and provides feedback adjustment parameters based on the interface.

[0006] Furthermore, the operation process of the intelligent diagnosis module includes the following steps: Performing time series segmentation and feature fusion on the operation data, and constructing a multimodal feature vector based on a time-frequency domain analysis method; Based on the multimodal feature vector, a generative adversarial network is used, wherein the generator simulates the potential fault state of the device by perturbing the collected data, and the discriminator performs dynamic adversarial optimization on normal and abnormal data to generate fault-sensitive features; Based on the fault-sensitive characteristics, according to the actual spatial distribution and operating status of the substation equipment, the power equipment is mapped into a graph structure, where the graph nodes represent a single power equipment and the graph edges represent the electrical or physical associations between the equipment, thereby generating a regionalized graph structure; Based on the regionalized graph structure, the graph neural network algorithm is used to analyze the correlation characteristics and fault propagation patterns between power equipment, and identify abnormal nodes and abnormal paths; Based on the results of abnormal pattern analysis, the abnormal types and impact ranges in the equipment operation status are classified and identified, and a power equipment diagnostic report is generated.

[0007] Furthermore, the fault-sensitive characteristics include frequency characteristics, vibration modes, electrical parameters and environmental coupling characteristics of the power equipment.

[0008] Furthermore, the operation process of the risk assessment module includes the following steps: Based on the power equipment diagnostic report, a mapping relationship between abnormal patterns and potential risk factors is established to extract risk factor features; Based on the risk factor characteristics, a fractional-order differential model is used to perform high-precision dynamic modeling of the equipment operation status, capture the rate and trend of power equipment state changes, and generate a power equipment operation status characteristic curve; Based on the power equipment operating status characteristic curve, the risk assessment index is dynamically optimized through the reinforcement learning algorithm to build a multi-level risk assessment framework including operation reliability, failure probability and risk propagation degree; Based on the multi-level risk assessment framework and the power equipment diagnostic report, dynamically analyze the current operating status and potential risks of the equipment to generate an equipment risk level; Based on the dynamic risk analysis results, a real-time risk assessment curve is generated.

[0009] Furthermore, the formula of the fractional-order differential model is as follows:

[0010] in, represents the fractional derivative of the power equipment operating state x(t), that is, the rate of change of the power equipment operating state; x(t) represents the operating state characteristic value of the power equipment; t represents the current time; q and represents the fractional order, q is the dominant order, is the fractional order corresponding to other characteristic variables; Indicates the contribution of different operating data to the dynamic change of the overall state; Indicates the disturbance intensity of environmental fluctuations on the operating status of the equipment; Indicates the time decay rate of environmental influence; Indicates the response strength of the equipment operation status to different time delays; Indicates the power equipment at the time point The external input signal received; Represents uncontrollable random factors.

[0011] Furthermore, the formula of the reinforcement learning algorithm is as follows:

[0012] in, Indicates the effect of optimizing risk assessment of power equipment; is the state parameter at time step t; P represents the update weight, which controls the ratio of historical Q value to current optimization benefit; Indicates measurement action Direct contribution to the optimization of the current risk assessment; D represents the discount factor; M represents the number of multi-level risk assessment indicators; represents the weight of the risk assessment indicator at the i-th level; Indicates the next state In Action Optimization effect of the i-th layer indicator.

[0013] Furthermore, the collaborative early warning module adopts an edge node grading mechanism to dynamically optimize the allocation of collaborative computing tasks based on the geographical distribution characteristics and communication delays of the equipment, and constructs a node priority queue to balance the task load according to the node's computing power, network stability and the operational importance of regional equipment. At the same time, it combines the regional risk level to achieve real-time response of high-priority nodes.

[0014] Furthermore, the interactive control module adopts a multi-objective decision-making optimization algorithm based on fuzzy neural networks to adaptively analyze the equipment operating status and risk assessment results, generate an optimized control strategy, and dynamically adjust the strategy execution path through a visual interface.

[0015] On the other hand, a monitoring and early warning method for substation power equipment includes the following steps: Collect operating data of power equipment in substations; Based on the operation data, a generative adversarial network combined with a graph neural network algorithm is used to perform in-depth analysis and abnormal pattern recognition on the operation characteristics of the power equipment to generate a power equipment diagnosis report; Based on the power equipment diagnostic report, a real-time risk assessment curve is generated based on a dynamic assessment framework combining a fractional-order differential model and a reinforcement learning algorithm; By building a collaborative architecture based on edge computing, the real-time risk assessment curve is processed in a multi-node collaborative manner, and a graded warning plan is generated according to the area where the equipment is located and the operating level; The adaptive optimization algorithm based on fuzzy neural network displays the equipment operation status, fault information and early warning suggestions through the human-computer interaction interface, and provides feedback adjustment parameters based on the interface.

[0016] The beneficial effects of the present invention are: The data acquisition module in the present invention can fully obtain the key data of the operation of power equipment, providing a high-quality data basis for subsequent analysis; the intelligent diagnosis module combined with the generative adversarial network and graph neural network algorithm performs in-depth analysis and abnormal pattern recognition on the operation data, which can quickly and accurately discover potential problems and improve the accuracy of diagnosis. Through the risk assessment module, the fractional differential model is used to accurately model the rate and trend of equipment state change, and the risk assessment index is dynamically optimized by reinforcement learning, so as to realize the dynamic and real-time adjustment of the risk level, generate a high-precision risk assessment curve, and provide effective support for the safety of equipment operation. The collaborative early warning module builds a multi-node collaborative architecture based on edge computing, which can efficiently distribute the risk assessment results, and generate a graded early warning scheme according to the equipment area and operation level, which improves the response speed and pertinence of the early warning, and improves the reliability and expansion capability of the system. The interactive control module uses the adaptive optimization algorithm of the fuzzy neural network, combined with an intuitive human-computer interaction interface, to display the equipment operation status, fault information and early warning suggestions in real time, and supports users to adjust parameters through feedback, thereby realizing real-time optimization of the operation status and improving the ease of use and controllability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a module schematic diagram of a monitoring and early warning system for power equipment in a substation of the present invention.

[0018] Figure 2 It is a flowchart of the operation process of the risk assessment module provided by an embodiment of the present invention.

[0019] Figure 3 The present invention is a flow chart of a monitoring and early warning method for power equipment in a substation. DETAILED DESCRIPTION

[0020] See also Figure 1-Figure 3 As shown, the present invention relates to a monitoring and early warning method and system for power equipment in a substation.

[0021] Example 1 A monitoring and early warning system for power equipment in a substation, comprising a data acquisition module, an intelligent diagnosis module, a risk assessment module, a collaborative early warning module and an interactive control module which are sequentially connected in communication; The data acquisition module is used to collect the operating data of the power equipment in the substation, and the operating data includes the current, voltage, temperature, humidity and vibration signals of the power equipment; Specifically, by deploying various types of sensors and collection devices, key operating parameters of substation power equipment are monitored in real time, including current, voltage, temperature, humidity and vibration signals.

[0022] Current and voltage acquisition: By installing high-precision current transformers and voltage transformers, accurate acquisition of equipment electrical parameters is achieved, and noise in the collected data is eliminated through filtering and anti-interference algorithms.

[0023] Temperature and humidity collection: Multiple temperature and humidity sensors are arranged inside and around the equipment, and a distributed collection solution is adopted to monitor the temperature and humidity change trends of the equipment environment.

[0024] Vibration signal acquisition: By installing an accelerometer or MEMS sensor, the vibration signal of the equipment is obtained, and the vibration spectrum characteristics are extracted using the FFT (Fast Fourier Transform) algorithm to determine whether there is any mechanical abnormality in the equipment.

[0025] Communication interface: The data acquisition module integrates multiple communication protocols (such as Modbus, IEC 61850, MQTT), and performs local data storage and preliminary processing through edge computing devices to ensure the real-time and integrity of data transmission.

[0026] The intelligent diagnosis module is used to generate a diagnosis report for the power equipment by using a generative adversarial network combined with a graph neural network algorithm to perform in-depth analysis and abnormal pattern recognition on the operation characteristics of the power equipment based on the operation data; The operation process of the intelligent diagnosis module includes the following steps: Performing time series segmentation and feature fusion on the operation data, and constructing a multimodal feature vector based on a time-frequency domain analysis method; Based on the multimodal feature vector, a generative adversarial network is used, wherein the generator simulates the potential fault state of the equipment by perturbing the collected data, and the discriminator performs dynamic adversarial optimization on normal and abnormal data to generate fault-sensitive features; the fault-sensitive features include frequency characteristics, vibration modes, electrical parameters and environmental coupling characteristics of the power equipment; Specifically, a multimodal feature vector is input, and the feature data is perturbed through the generator of the Generative Adversarial Network (GAN) to simulate the potential fault state of the equipment. The generator uses a structure that combines a fully connected neural network with a convolutional neural network (CNN) to generate feature samples containing potential fault characteristics. The discriminator learns to distinguish between normal data and abnormal data by inputting the generated feature samples and the actual collected operating data. The discriminator dynamically adjusts the weights, enhances the extraction of abnormal features, and suppresses the invalid perturbations of the generator.

[0027] Compare the features of the generated abnormal data and normal data and extract sensitive features, including: Frequency characteristics: harmonic frequency changes and harmonic distortion during equipment operation.

[0028] Vibration Mode: The variation in the vibration amplitude and frequency spectrum of a mechanical component.

[0029] Electrical parameters: current and voltage fluctuation characteristics and overload phenomena.

[0030] Environmental coupling characteristics: changes in the correlation between equipment operation and environmental humidity and temperature.

[0031] Based on the fault-sensitive characteristics, according to the actual spatial distribution and operating status of the substation equipment, the power equipment is mapped into a graph structure, where the graph nodes represent a single power equipment and the graph edges represent the electrical or physical associations between the equipment, thereby generating a regionalized graph structure; Specifically, multiple power equipment are mapped into a regionalized graph structure according to the spatial distribution and operating status of the equipment: Node definition: Each power device corresponds to a node, and the node attributes include the fault-sensitive characteristics of the device.

[0032] Edge definition: Based on the electrical network or physical structure, define the associated edges between devices (such as current coupling, mechanical connection).

[0033] The strength of association between devices is described by a weighted adjacency matrix, where the weights are determined by the distance between devices, transmission power, and fault correlation. For example, the electrical coupling edge between transformers and switches has a higher weight, while the physical connection edge between distant devices has a lower weight. The global graph is divided into multiple regional subgraphs based on the geographical distribution or functional zoning of the devices (such as the main transformer area and the high-voltage switch area). Subgraph processing improves computational efficiency and reduces interference from edge devices.

[0034] Based on the regionalized graph structure, the graph neural network algorithm is used to analyze the correlation characteristics and fault propagation patterns between power equipment, and identify abnormal nodes and abnormal paths; Specifically, the graph structure is modeled using a graph convolutional network (GCN): neighbor node features are aggregated through the graph convolution layer to update the global properties of the nodes. The weights of the graph edges are dynamically adjusted in combination with fault-sensitive features to enhance the significance of the abnormal propagation path. The propagation path of the fault in the graph structure is simulated to analyze the possible impact range. High-risk nodes (abnormal equipment) and key edges (connections that may cause cascading failures) are identified. Through the output layer of GCN, the operating status of each node is classified and abnormal nodes are marked. The critical path of abnormal propagation is located based on the shortest path algorithm or random walk method based on the graph.

[0035] Based on the results of abnormal pattern analysis, the abnormal types and impact ranges in the equipment operation status are classified and identified, and a power equipment diagnostic report is generated.

[0036] It should be noted that the analysis results of the generative adversarial network and the graph neural network are combined to classify the operating status of abnormal nodes: Electrical anomalies: such as current overload and voltage fluctuation.

[0037] Mechanical abnormalities: such as excessive vibration and running jams.

[0038] Abnormal environment: such as excessive humidity leading to reduced insulation.

[0039] Based on the analysis results of the fault propagation path, evaluate the range of equipment that the anomaly may affect. Mark key equipment and areas to determine the priority order. Output a diagnostic report, including an overview of the equipment status, anomaly type, impact range, fault cause analysis, and repair suggestions. Provide visual diagnostic results, such as equipment operation status distribution diagram, anomaly propagation path diagram, etc.

[0040] The risk assessment module is used to perform high-precision modeling of the rate of change and trend of the operating state of the power equipment based on the power equipment diagnostic report and a dynamic assessment framework combining a fractional-order differential model with a reinforcement learning algorithm, and dynamically optimize the equipment risk level assessment index using reinforcement learning to generate a real-time risk assessment curve; The operation process of the risk assessment module includes the following steps: Based on the power equipment diagnostic report, a mapping relationship between abnormal patterns and potential risk factors is established to extract risk factor features; Specifically, key data such as historical operation data, abnormal records, and fault reports are obtained from the power equipment diagnostic report. These data are preprocessed, including noise removal, missing value filling, and normalization. Clustering algorithms are used to classify patterns of abnormal data and mark different types of abnormalities (such as temperature abnormalities, current abnormalities, or vibration abnormalities). Mapping with potential risk factors is established based on the type of abnormality, for example: temperature abnormalities are associated with thermal failures, current abnormalities are associated with overloads, and vibration abnormalities are associated with mechanical failure risks. Feature indicators corresponding to each abnormality are extracted, such as average temperature, temperature rise rate, current fluctuation amplitude, etc. A risk factor feature matrix is ​​constructed, which contains time series information, feature statistics, and importance weights to provide input for subsequent dynamic modeling.

[0041] Based on the risk factor characteristics, a fractional-order differential model is used to perform high-precision dynamic modeling of the equipment operation status, capture the rate and trend of power equipment state changes, and generate a power equipment operation status characteristic curve; It should be noted that key input variables (such as temperature rise rate, vibration amplitude, load change amplitude) are selected from the risk factor matrix. These variables should have high time correlation and physical meaning. The real-time data is segmented using sliding window technology (for example, the window length is set to 5 minutes), and fractional differential modeling is performed on each segment of data. The parameters fitted within the window include trend (such as rise / fall), rate (amount of change per unit time), etc. The rate and trend of equipment state change are captured by the slope and curvature of the fractional differential output curve. For example: a positive slope and an increasing curvature of the curve indicate that the equipment state is rapidly deteriorating; a negative slope and a decreasing curvature of the curve indicate that the state has returned to normal. The fractional order q is optimized using gradient descent or genetic algorithm to ensure that the curve fitting error is minimized. Verify the curve fitting accuracy. Visualize the result curve of fractional differential modeling and mark key points (such as risk trigger points and change acceleration points) with different colors. Combined with the diagnostic report, annotate the operating trend and potential risk points of the equipment.

[0042] Based on the power equipment operating status characteristic curve, the risk assessment index is dynamically optimized through the reinforcement learning algorithm to build a multi-level risk assessment framework including operation reliability, failure probability and risk propagation degree; Specifically, define the initial risk assessment indicator set, including operational reliability (such as MTTF / MTBF), failure probability (such as historical data statistics), and risk propagation degree (such as the coefficient of mutual influence between devices). Set the initial weight value of each indicator. Use reinforcement learning algorithms (such as deep Q learning or Actor-Critic algorithm) and use equipment historical data as a training set to build a risk assessment indicator optimization model. Through the dynamic relationship between the equipment operating status characteristic curve and risk factors, set a reward and punishment mechanism (such as giving positive rewards for reducing risk levels and giving positive rewards for improving the accuracy of predicting failures). According to the optimal weight output by reinforcement learning, optimize the weight distribution of each risk assessment indicator in the framework. Comprehensively analyze the operational reliability, failure probability, and risk propagation degree according to the hierarchical structure (equipment level, system level, network level) to form a multi-level evaluation framework.

[0043] Based on the multi-level risk assessment framework and the power equipment diagnostic report, dynamically analyze the current operating status and potential risks of the equipment to generate an equipment risk level; Based on the dynamic risk analysis results, a real-time risk assessment curve is generated.

[0044] Furthermore, the formula of the fractional-order differential model is as follows:

[0045] in, represents the fractional derivative of the power equipment operating state x(t), that is, the rate of change of the power equipment operating state; x(t) represents the operating state characteristic value of the power equipment; t represents the current time; q and represents the fractional order, q is the dominant order, is the fractional order corresponding to other characteristic variables; Indicates the contribution of different operating data to the dynamic change of the overall state; Indicates the disturbance intensity of environmental fluctuations on the operating status of the equipment; Indicates the time decay rate of environmental influence; Indicates the response strength of the equipment operation status to different time delays; Indicates the power equipment at the time point The external input signal received; Represents uncontrollable random factors.

[0046] Kernel Function The specific formula is as follows:

[0047] in, Represents the Gamma function, used for normalization; Indicates the memory degree parameter, which indicates the sensitivity of the device to the past operating status; when When it is larger, the device is more sensitive to the recent historical operating status (such as the most recent current or vibration signal); When it is smaller, the device has a stronger cumulative memory of more distant historical operating data (such as temperature fluctuations and humidity effects).

[0048] The formula of the reinforcement learning algorithm is as follows:

[0049] in, Indicates the effect of optimizing risk assessment of power equipment; is the state parameter at time step t; P represents the update weight, which controls the ratio of historical Q value to current optimization benefit, and its value range is ; Indicates measurement action Direct contribution to the optimization of the current risk assessment; D represents the discount factor; M represents the number of multi-level risk assessment indicators, such as equipment operation stability, failure probability, and risk propagation range; represents the weight of the risk assessment indicator at the i-th level; Indicates the next state In Action Under this condition, the optimization effect of the i-th layer indicator is as follows:

[0050] in, represents the optimization increment of the i-th level indicator; Indicates the maximum benchmark value of the i-th level indicator; represents the risk adjustment coefficient of the indicator at the i-th level, which is used to adjust the sensitivity of the indicator to the risk assessment results; Indicates the risk value corresponding to the i-th level indicator, which changes dynamically according to the equipment operation status.

[0051] The collaborative warning module is used to construct a collaborative architecture based on edge computing, perform multi-node collaborative processing on the real-time risk assessment curve, and generate a graded warning plan according to the area where the equipment is located and the operating level; the collaborative warning module uses an edge node grading mechanism based on the geographical distribution characteristics and communication delays of the equipment to dynamically optimize the allocation of collaborative computing tasks, and by constructing a node priority queue, balance the task load according to the computing power of the node, network stability and the operating importance of the regional equipment, and at the same time realize real-time response of high-priority nodes in combination with the regional risk level.

[0052] In one embodiment, multiple edge computing nodes (such as high-performance edge gateways or edge servers) are deployed in the substation area according to the distribution characteristics of power equipment. These nodes have independent computing capabilities, can receive data, analyze risks, and generate local warnings. Edge nodes are interconnected through optical fiber communication, 5G network or wireless private network to achieve low-latency and highly reliable data exchange. Each edge node performs preliminary processing on the risk assessment curve of the equipment it is responsible for and generates preliminary warning data. Through real-time communication between edge nodes, cross-node data is integrated, the overall risk trend of the region is analyzed, and an optimized global warning plan is generated. Distributed storage technology is used to ensure data consistency. Even if a node fails, other nodes can quickly take over its tasks to avoid information loss.

[0053] Task levels (such as emergency, high priority, and general) are divided according to the operating status of the equipment and the risk level of the area. Emergency tasks include real-time response to high-risk equipment; general tasks are routine monitoring of low-risk equipment. Each node maintains a task queue and prioritizes tasks for high-risk equipment or important areas. Task priorities are dynamically updated based on the geographical location of the equipment, operational importance, and communication delay. Edge nodes dynamically adjust task allocation based on computing power and real-time network conditions to ensure efficient use of computing resources. An allocation mechanism based on load prediction is introduced to avoid node overload or resource waste. A reinforcement learning algorithm is used to adaptively optimize the task allocation strategy, and the efficiency of task allocation is improved by continuously learning the performance of edge nodes and the execution effect of tasks.

[0054] The warning level is generated based on indicators such as power equipment health index, real-time risk assessment value, and regional equipment importance. The specific classification is as follows: High-priority warning: Generate real-time warnings for important equipment (such as main transformers, switchgear) and equipment with higher operating risk levels, including: fault type prediction and recommended emergency response strategies.

[0055] Medium priority warning: Provides trend forecasts and maintenance recommendations for equipment with fluctuating operating conditions.

[0056] Low priority warning: For equipment that is operating normally, only monitoring data is recorded and a regular operation report is generated.

[0057] Generate diversified response plans for different warning levels: High-priority warnings directly notify the operation and maintenance personnel and link the control module for rapid adjustment; medium and low priority warnings are stored as reference information for subsequent equipment optimization or regular maintenance; comprehensive analysis of multiple equipment risks in the region is conducted to assess the overall regional risk level. A regional-level warning plan is generated to guide the linkage management of equipment in the region.

[0058] Node levels are divided according to the performance (such as CPU, memory), network latency and reliability of edge nodes. High-performance nodes (such as servers with stronger computing power) prioritize complex tasks, and low-performance nodes are responsible for basic tasks. Dynamically adjust node priorities based on real-time device status changes and regional risk levels. Build a real-time priority adjustment algorithm to optimize task allocation based on the real-time status of edge nodes. Set up a backup node for each critical task. When the main node fails, the backup node immediately takes over the task to ensure the continuity of the early warning service. When the risk level of a substation area increases, the edge nodes in the surrounding areas can collaboratively support its computing tasks to form a cross-regional collaborative early warning network.

[0059] The generated warning information is pushed to the interactive control module for operation and maintenance personnel to view and provide feedback. Operation and maintenance personnel can confirm or adjust the warning suggestions through the interface, and the system will use the feedback results for model optimization. Regularly evaluate the accuracy of warnings, response time, and task allocation efficiency. Continuously optimize the warning algorithm through reinforcement learning to improve system performance.

[0060] The interactive control module is used for an adaptive optimization algorithm based on a fuzzy neural network, which displays the equipment operating status, fault information and early warning suggestions through a human-computer interaction interface, and provides feedback adjustment parameters based on the interface; the interactive control module adopts a multi-objective decision-making optimization algorithm based on a fuzzy neural network, adaptively analyzes the equipment operating status and risk assessment results, generates an optimized control strategy, and dynamically adjusts the strategy execution path through a visual interface.

[0061] In one embodiment, dynamic charts (such as line charts, bar charts, and heat maps) are used to display the operating status of the equipment, including core parameters such as current, voltage, temperature and humidity, and vibration signals. The risk assessment curve is dynamically updated, and the risk level is marked with different colors (such as green for low risk, yellow for medium risk, and red for high risk). The health summary of all monitored devices is displayed on the main interface, including the equipment operating status (healthy, abnormal, faulty), risk assessment value (expressed as health index HI), and warning level. Users can click on a specific device to enter the detailed page to view diagnostic reports and suggestions. The operating status and geographical distribution of each device are intuitively displayed through a 3D substation equipment model (digital twin). Warning information is superimposed on the equipment model in real time (such as areas with excessively high temperatures are highlighted in red).

[0062] The equipment status is evaluated by classification based on the fuzzy rule base (e.g., “if the temperature is high and the vibration is large, the health index is low”).

[0063] Fuzzy logic rule definition: Input variables: operating parameters (such as current, voltage, temperature, vibration) and risk assessment values.

[0064] Output variables: optimized control strategies (such as adjusting operating parameters or recommending early warning plans).

[0065] Use neural networks to adaptively adjust the weights of fuzzy rules and dynamically optimize decision results. Make trade-offs between multiple goals (such as operating efficiency, maintenance costs, and failure risks): prioritize stopping the operation of high-risk equipment; recommend adjusting operating parameters (such as reducing load) for medium-risk equipment; and maintain normal operation of low-risk equipment and reduce intervention.

[0066] The optimization control strategy generation process includes the following steps: Obtain the current equipment operating status and risk assessment data. Generate preliminary optimization suggestions based on the output of the fuzzy neural network and the actual equipment operating parameters. Input the preliminary optimization suggestions into the multi-objective decision model. Optimize the execution path in combination with the operating goals (such as reducing risks, extending equipment life, etc.). For example: Optimization path 1: Reduce equipment load → adjust the operating time period → enhance heat dissipation. Optimization path 2: Increase the equipment tolerance threshold → adjust the maintenance plan. Use simulation technology to predict the execution effect of the optimization strategy (such as reducing temperature, reducing vibration, etc.). The interface displays the prediction effect curves of different strategies to facilitate users to select the best solution.

[0067] Example 2 A monitoring and early warning method for substation power equipment comprises the following steps: Collect operating data of power equipment in substations; Based on the operation data, a generative adversarial network combined with a graph neural network algorithm is used to perform in-depth analysis and abnormal pattern recognition on the operation characteristics of the power equipment to generate a power equipment diagnosis report; Based on the power equipment diagnostic report, a real-time risk assessment curve is generated based on a dynamic assessment framework combining a fractional-order differential model and a reinforcement learning algorithm; By building a collaborative architecture based on edge computing, the real-time risk assessment curve is processed in a multi-node collaborative manner, and a graded warning plan is generated according to the area where the equipment is located and the operating level; The adaptive optimization algorithm based on fuzzy neural network displays the equipment operation status, fault information and early warning suggestions through the human-computer interaction interface, and provides feedback adjustment parameters based on the interface.

[0068] In this embodiment, a monitoring and early warning method for power equipment in a substation is applied to a monitoring and early warning system for power equipment in a substation described in Example 1, which will not be described in detail here.

[0069] In summary, the present invention realizes high-precision acquisition of key operating parameters such as current, voltage, temperature and humidity, and vibration signals through multiple types of sensors (such as current transformers, voltage transformers, temperature and humidity sensors, and acceleration sensors), and combines filtering and anti-interference algorithms to ensure the accuracy and integrity of data, providing a reliable basis for subsequent analysis. The algorithm that combines generative adversarial networks (GANs) and graph neural networks (GNNs) can not only simulate potential fault states, but also identify the complex correlation characteristics and fault propagation modes of power equipment. Through multimodal feature fusion and dynamic optimization, the accuracy and comprehensiveness of fault diagnosis are improved.

[0070] The present invention uses a fractional differential model to perform high-precision modeling on the rate of change and trend of the equipment's operating status, capturing the dynamic characteristics of potential risks. The risk assessment indicators are dynamically optimized by combining a reinforcement learning algorithm to generate a real-time risk assessment curve to ensure real-time control of the equipment's health status. Multi-node collaborative processing is achieved through an edge computing architecture, task allocation is dynamically optimized, and communication delays are reduced. A hierarchical warning scheme is generated based on regional risk levels and equipment importance to ensure that high-risk equipment responds first, improving overall warning efficiency and reliability.

[0071] The present invention displays the equipment operation status, fault information and risk curve through a visual interface, providing clear and intuitive equipment health information. At the same time, combined with the fuzzy neural network adaptive optimization algorithm, users can adjust system parameters in real time to improve decision-making efficiency and optimize operation strategies. The system optimizes the utilization of computing resources through task load prediction and dynamic allocation mechanism, and introduces a backup node mechanism to ensure the continuous operation capability of the system, and can maintain stable early warning services even when a node fails.

[0072] According to the equipment operation status and diagnostic reports, the system classifies and analyzes the abnormality types (such as electrical, mechanical, and environmental abnormalities) and provides detailed fault causes and impact ranges. Early warning schemes cover different levels, from emergency tasks to routine monitoring, to ensure flexible response in different risk scenarios. Through early warning of high-risk equipment, optimization of operating parameters and fault response strategies, the system can effectively reduce equipment operation risks, extend equipment service life, and reduce maintenance costs and the probability of operation interruption.

[0073] The above implementation modes are merely descriptions of the preferred implementation modes of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary engineering and technical personnel in the field shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A monitoring and early warning system for power equipment in a substation, characterized in that: It includes a data acquisition module, an intelligent diagnosis module, a risk assessment module, a collaborative early warning module and an interactive control module which are sequentially connected in communication; The data acquisition module is used to collect the operating data of the power equipment in the substation, and the operating data includes the current, voltage, temperature, humidity and vibration signals of the power equipment; The intelligent diagnosis module is used to generate a diagnosis report for the power equipment by using a generative adversarial network combined with a graph neural network algorithm to perform in-depth analysis and abnormal pattern recognition on the operation characteristics of the power equipment based on the operation data; The risk assessment module is used to perform high-precision modeling of the rate of change and trend of the operating state of the power equipment based on the power equipment diagnostic report and a dynamic assessment framework combining a fractional-order differential model with a reinforcement learning algorithm, and dynamically optimize the equipment risk level assessment index using reinforcement learning to generate a real-time risk assessment curve; The collaborative warning module is used to perform multi-node collaborative processing on the real-time risk assessment curve by building a collaborative architecture based on edge computing, and generate a graded warning plan according to the area where the equipment is located and the operation level; The interactive control module is used for adaptive optimization algorithm based on fuzzy neural network, displays equipment operation status, fault information and early warning suggestions through human-computer interaction interface, and provides feedback adjustment parameters based on the interface.

2. A monitoring and early warning system for power equipment in a substation according to claim 1, characterized in that: The operation process of the intelligent diagnosis module includes the following steps: Performing time series segmentation and feature fusion on the operation data, and constructing a multimodal feature vector based on a time-frequency domain analysis method; Based on the multimodal feature vector, a generative adversarial network is used, wherein the generator simulates the potential fault state of the device by perturbing the collected data, and the discriminator performs dynamic adversarial optimization on normal and abnormal data to generate fault-sensitive features; Based on the fault-sensitive characteristics, according to the actual spatial distribution and operating status of the substation equipment, the power equipment is mapped into a graph structure, where the graph nodes represent a single power equipment and the graph edges represent the electrical or physical associations between the equipment, thereby generating a regionalized graph structure; Based on the regionalized graph structure, the graph neural network algorithm is used to analyze the correlation characteristics and fault propagation patterns between power equipment, and identify abnormal nodes and abnormal paths; Based on the results of abnormal pattern analysis, the abnormal types and impact ranges in the equipment operation status are classified and identified, and a power equipment diagnostic report is generated.

3. A monitoring and early warning system for power equipment in a substation according to claim 2, characterized in that: The fault-sensitive characteristics include frequency characteristics, vibration modes, electrical parameters and environmental coupling characteristics of the power equipment.

4. A monitoring and early warning system for power equipment in a substation according to claim 1, characterized in that: The operation process of the risk assessment module includes the following steps: Based on the power equipment diagnostic report, a mapping relationship between abnormal patterns and potential risk factors is established to extract risk factor features; Based on the risk factor characteristics, a fractional-order differential model is used to perform high-precision dynamic modeling of the equipment operation status, capture the rate and trend of power equipment state changes, and generate a power equipment operation status characteristic curve; Based on the power equipment operating status characteristic curve, the risk assessment index is dynamically optimized through the reinforcement learning algorithm to build a multi-level risk assessment framework including operation reliability, failure probability and risk propagation degree; Based on the multi-level risk assessment framework and the power equipment diagnostic report, dynamically analyze the current operating status and potential risks of the equipment to generate an equipment risk level; Based on the dynamic risk analysis results, a real-time risk assessment curve is generated.

5. A monitoring and early warning system for power equipment in a substation according to claim 4, characterized in that: The formula of the fractional differential model is as follows: ； in, represents the fractional derivative of the power equipment operating state x(t), that is, the rate of change of the power equipment operating state; x(t) represents the operating state characteristic value of the power equipment; t represents the current time; q and represents the fractional order, q is the dominant order, is the fractional order corresponding to other characteristic variables; Indicates the contribution of different operating data to the dynamic change of the overall state; Indicates the disturbance intensity of environmental fluctuations on the operating status of the equipment; Indicates the time decay rate of environmental influence; Indicates the response strength of the equipment operation status to different time delays; Indicates the power equipment at the time point The external input signal received; Represents uncontrollable random factors.

6. A monitoring and early warning system for power equipment in a substation according to claim 4, characterized in that: The formula of the reinforcement learning algorithm is as follows: ； in, Indicates the effect of optimizing risk assessment of power equipment; is the state parameter at time step t; P represents the update weight, which controls the ratio of historical Q value to current optimization benefit; Indicates measurement action Direct contribution to the optimization of the current risk assessment; D represents the discount factor; M represents the number of multi-level risk assessment indicators; represents the weight of the risk assessment indicator at the i-th level; Indicates the next state In Action Optimization effect of the i-th layer indicator.

7. A monitoring and early warning system for power equipment in a substation according to claim 1, characterized in that: The collaborative early warning module adopts an edge node grading mechanism to dynamically optimize the allocation of collaborative computing tasks based on the geographical distribution characteristics and communication delays of the equipment, and constructs a node priority queue to balance the task load according to the computing power of the node, network stability and the operational importance of the regional equipment. At the same time, it combines the regional risk level to achieve real-time response of high-priority nodes.

8. A monitoring and early warning system for power equipment in a substation according to claim 1, characterized in that: The interactive control module adopts a multi-objective decision-making optimization algorithm based on fuzzy neural networks to adaptively analyze the equipment operation status and risk assessment results, generate an optimized control strategy, and dynamically adjust the strategy execution path through a visual interface.

9. A monitoring and early warning method for power equipment in a substation, characterized in that: The method is applied to a monitoring and early warning system for power equipment in a substation as claimed in any one of claims 1 to 8, comprising the following steps: Collect operating data of power equipment in substations; Based on the operation data, a generative adversarial network combined with a graph neural network algorithm is used to perform in-depth analysis and abnormal pattern recognition on the operation characteristics of the power equipment to generate a power equipment diagnosis report; Based on the power equipment diagnostic report, a real-time risk assessment curve is generated based on a dynamic assessment framework combining a fractional-order differential model and a reinforcement learning algorithm; By building a collaborative architecture based on edge computing, the real-time risk assessment curve is processed in a multi-node collaborative manner, and a graded warning plan is generated according to the area where the equipment is located and the operating level; The adaptive optimization algorithm based on fuzzy neural network displays the equipment operation status, fault information and early warning suggestions through the human-computer interaction interface, and provides feedback adjustment parameters based on the interface.

Images