Method and system for identifying and early warning defect abnormity of power grid equipment in advance

Abstract

The invention discloses a method and a system for identifying and early warning defects and abnormities of power grid equipment in advance, and the method comprises the steps: collecting the operation data of the power grid equipment, including historical data and real-time data, and carrying out the data cleaning and preprocessing; based on the preprocessed operation data, establishing a digital twinborn model of the power grid equipment, and updating the model in real time by adopting a Kalman filtering technology; a finite element analysis module is integrated in the digital twin model, stress distribution of equipment under different operation conditions is simulated, and potential fault points are identified; performing anomaly detection on the equipment operation data based on an isolated forest algorithm, and generating an early warning signal when an anomaly is detected; according to the method, the fault abnormity problem of the power grid equipment can be comprehensively predicted in advance from multiple angles, the accuracy and efficiency of early warning are improved, the safety and stability of the power grid are remarkably enhanced, and the occurrence of equipment faults is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of abnormality identification and detection of power grid equipment, and in particular relates to a method and system for early identification and early warning of abnormal defects in power grid equipment. Background Art

[0002] With the rapid development of the electric power industry, the scale of power grid systems continues to expand, their structures become increasingly complex, and the number and variety of grid equipment has increased significantly. Key equipment such as transformers and circuit breakers play a vital role in power transmission and distribution, and their operating status directly impacts the stability and reliability of the grid. In recent years, the widespread adoption of smart grid technology and IoT devices has provided technical support for real-time monitoring and data collection of grid equipment, driving the transformation of power systems towards digitalization and intelligence. At the same time, the power industry is experiencing a growing demand for equipment status monitoring and fault warnings to ensure safe grid operation and improve operational efficiency.

[0003] Currently, power grid equipment fault early warning technology has made some progress. Patent CN115549094A proposes a smart grid substation equipment early warning assessment method. By acquiring substation equipment information, establishing substation working relationships, and combining equipment monitoring data for status assessment and fault diagnosis, it achieves dynamic analysis and timely early warning. Patent CN106529696A uses equipment data level values ​​and a weight judgment matrix to calculate equipment risk impact and generate early warning information based on preset thresholds.

[0004] However, in complex, dynamically changing, large-scale power grid environments, traditional methods may lack adaptability and accuracy, struggling to cope with the combined impact of multiple factors such as load fluctuations, environmental disturbances, and equipment aging. Furthermore, some early warning models rely on fixed thresholds or static weights and lack dynamic optimization of real-time data, resulting in high false alarm rates or delayed warnings. Summary of the Invention

[0005] Purpose of the invention: The purpose of the present invention is to provide a method for early identification and early warning of power grid equipment defects and anomalies that meets the needs of high-precision and high-reliability intelligent early warning, optimizes algorithms and improves the robustness of the system; on the other hand, to provide an early identification and early warning system for power grid equipment defects and anomalies.

[0006] Technical solution: The early identification and warning method of the present invention includes the following steps:

[0007] (1) Collecting operational data of power grid equipment, including historical and real-time data, and performing data cleaning and preprocessing, building a complete and reliable data base for equipment operation, solving the problems of incomplete data and high noise interference in traditional monitoring systems, and providing high-quality data input for subsequent analysis. This step significantly improves data availability and analysis accuracy;

[0008] (2) Based on the pre-processed operating data, a digital twin model of the power grid equipment is established, and the Kalman filter technology is used to update the model in real time, realizing dynamic simulation and precise tracking of the operating status of the power grid equipment. This breaks through the limitation of traditional static models that are difficult to reflect the real-time status of the equipment, and enables the system to have the ability to predict the operating trends of the equipment;

[0009] (3) Integrating a finite element analysis module into the digital twin model simulates the stress distribution of the equipment under different operating conditions, identifies potential fault points, and achieves cross-dimensional fault prediction from electrical parameters to mechanical status, significantly improving the depth and breadth of fault detection;

[0010] (4) Anomaly detection is performed on equipment operation data based on the isolation forest algorithm. When an anomaly is detected, an early warning signal is generated, which effectively solves the industry problem of insufficient labeled data. The anomaly scoring mechanism is used to achieve sensitive detection of early weak faults. Its unique tree structure enables the system to have the advantage of processing high-dimensional data, significantly improving the timeliness and reliability of early warnings.

[0011] Preferably, the operating data in step 1 includes current, voltage, frequency, temperature, humidity, equipment health status data, environmental data and load data.

[0012] By collecting electrical parameters such as current, voltage, and frequency, and combining them with environmental data such as temperature and humidity, as well as equipment health status and load data, a multi-dimensional equipment operation monitoring system has been constructed, achieving comprehensive perception of the operating status of power grid equipment. This fusion collection method of multi-source heterogeneous data breaks through the limitations of traditional monitoring systems that only focus on a single electrical parameter. It can more comprehensively reflect the actual operating status of the equipment, provide a richer data foundation for subsequent fault warning analysis, and significantly improve the system's ability to identify complex fault modes.

[0013] Preferably, the establishment of the digital twin model in step 2 includes:

[0014] The grid topology is modeled as a graph theory model G = (N, E), where N represents the set of grid element nodes, E represents the set of connecting edges between elements, and the attributes of nodes and edges include resistance, reactance, and load parameters;

[0015] A power grid operation state model is established based on the power flow equation, which is expressed as:

[0016] P i -P j =V i ·V j ·(G ij ·cos(θ i -θ j )+Bij ·sin(θ i -θ j ))

[0017] Q i -Q j =V i ·V j ·(G ij ·sin(θ i -θ j )-B ij ·cos(θ i -θ j ))

[0018] Where, P i and Q i are the active and reactive power at node i, V i is the voltage at node i, θ i θ j is the phase angle between nodes i and j, G ij and B ij are the real and imaginary parts of the admittance matrix respectively.

[0019] By modeling the grid topology as a graph theory model G=(N,E) and establishing an operating status model based on the power flow equation, accurate digital mapping of the grid physical system is achieved; the electrical characteristics of the grid components are fully characterized by the attribute parameters of the nodes and edges (resistance, reactance and load), and combined with the mathematical description of key operating parameters such as voltage and power by the power flow equation, a digital twin model with physical reality is constructed. This modeling method not only accurately reflects the topology and dynamic behavior of the power grid, but also effectively simulates the energy flow and loss during power transmission, providing a high-precision simulation analysis basis for subsequent fault warning, and significantly improving the model's ability to characterize the grid operating status and prediction accuracy.

[0020] Preferably, the solution of the power flow equation adopts the Runge-Kutta numerical method to perform time stepping simulation, and the simulation process includes setting initial conditions of node voltage and power, and simulating the dynamic operation process of the power grid through iterative calculation.

[0021] By setting precise initial conditions and performing iterative calculations, high-precision simulation of the dynamic operation process of the power grid is achieved. This method has the characteristics of good numerical stability and high computational accuracy. It can effectively handle the nonlinear characteristics and complex dynamic behaviors in power grid operation, and overcomes the shortcomings of traditional solution methods in convergence and computational efficiency. Through time-stepping simulation, it can accurately capture the continuous change process of the power grid operation state, providing a reliable numerical calculation basis for the digital twin model, significantly improving the system's simulation capabilities of the power grid transient process and dynamic response, and making fault warnings more timely and accurate.

[0022] Preferably, the update of the digital twin model is implemented using a Kalman filter algorithm, and the calculation equation of the Kalman filter is:

[0023] x k =A·x k -1+B·u k +w k

[0024] Where u k is the state vector, A is represented by node voltage and power in the power grid; B is the state transfer matrix; u k is the control input matrix; w k For control input.

[0025] By using state-space equations to optimally estimate key parameters such as grid node voltage and power, the technical difficulty of traditional static models being difficult to adapt to real-time changes in the grid is effectively resolved. By integrating actual measurement data with model prediction values, high-precision tracking and prediction of the grid's operating status are achieved while taking into account process noise and observation noise. This significantly improves the real-time and accuracy of the digital twin model, enabling the system to capture subtle changes in equipment status in a timely manner, and providing a reliable dynamic analysis basis for early fault warning.

[0026] Preferably, the establishment of the finite element analysis module in step 3 includes:

[0027] Based on the equipment operation data output by the digital twin model, a finite element model that includes the equipment's geometric structure and material properties is constructed. Finite element analysis software ANSYS is used for meshing to analyze the stress conditions that cause equipment deformation or damage under different conditions. The geometric structure includes shape, size, and key characteristic parameters; the material properties include elastic modulus, Poisson's ratio, and thermal conductivity.

[0028] Based on the actual operating data of the equipment, mechanical load, electromagnetic force and thermal stress are set as boundary conditions. Finite element analysis is used to calculate the stress distribution of the equipment under various working conditions and identify potential failure points.

[0029] By deeply integrating finite element analysis modules with digital twin models, a technological breakthrough has been achieved in multi-physics coupled simulation analysis of power grid equipment. Based on precise geometric structure and material property modeling, combined with ANSYS software's advanced meshing technology, this system accurately simulates the stress distribution of equipment under the combined effects of mechanical loads, electromagnetic forces, and thermal stresses. By converting actual operating data into boundary conditions, the system dynamically analyzes the stress state of equipment under different operating conditions, accurately identifying stress concentration points and potential fault locations that are difficult to detect with traditional monitoring methods. This enables comprehensive fault prediction from electrical parameters to mechanical conditions, significantly improving the reliability and lead time of the early warning system.

[0030] Preferably, the anomaly detection of the equipment operation data based on the isolation forest algorithm in step 4 includes:

[0031] An isolation forest model consisting of multiple isolated trees is constructed, and the anomaly score S(x) is obtained by calculating the path length h(x) of the sample in the tree structure. The calculation formula is:

[0032]

[0033] Where E|h(x)| is the average path length of sample x in the isolation forest, c(ψ) is the normalization factor, and its calculation formula is:

[0034]

[0035] Where ψ is a key parameter of the power grid equipment status, reflecting the equipment operating status; H(ψ-1) is the Heaviside step function, used to determine whether ψ is greater than 2; n is a normalization constant that affects the impact of ψ on the result; c(ψ) is used to describe the anomaly detection function in the power grid equipment fault early warning system.

[0036] When the abnormality score S(x) is lower than the preset threshold, it is determined that the equipment is abnormal and an early warning signal is generated.

[0037] Through the integrated modeling of multiple isolated trees and the path length calculation mechanism, efficient unsupervised anomaly detection of power grid equipment operating data is achieved; by calculating the anomaly score of the sample and setting a dynamic threshold, the early abnormal state of the equipment can be accurately identified. Its unique tree structure and path length evaluation method enable the system to process high-dimensional data, while overcoming the traditional method's reliance on labeled data, significantly improving the sensitivity of fault warning (triggered when the anomaly score is lower than the threshold) and real-time performance, providing intelligent monitoring guarantee for the safe operation of power grid equipment.

[0038] The early identification and warning system of the present invention comprises:

[0039] Data acquisition and preprocessing module, used to collect operating data of power grid equipment, including historical data and real-time data, and perform data cleaning and preprocessing;

[0040] The digital twin modeling module is used to build a digital twin model of power grid equipment based on pre-processed operating data and update the model in real time using Kalman filtering technology;

[0041] Finite element analysis module, integrated into the digital twin model, is used to simulate the stress distribution of equipment under different operating conditions and identify potential failure points;

[0042] The anomaly detection and early warning module is used to detect anomalies in equipment operation data based on the isolation forest algorithm and generate an early warning signal when an anomaly is detected.

[0043] A computer-readable storage medium stores a computer program thereon, characterized in that when the computer program is executed by a processor, the method for early identification and warning of power grid equipment defects and abnormalities is implemented.

[0044] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for early identification and warning of defects and abnormalities in power grid equipment is implemented.

[0045] Beneficial effects: Compared with the existing technology, the present invention has the following significant advantages: 1. It can comprehensively and multi-angle predict the fault anomaly problems of power grid equipment in advance, improve the accuracy and efficiency of early warning, significantly enhance the security and stability of the power grid, and reduce the occurrence of equipment failures; 2. Through the dynamic update of the digital twin model and the application of Kalman filtering technology, real-time monitoring and accurate prediction of the status of power grid equipment are realized. Combined with the unsupervised learning characteristics of the isolation forest algorithm, the system can quickly identify anomalies without prior labels, greatly improving the timeliness and intelligence level of fault warning, so that operation and maintenance personnel can timely Take preventive measures; 3. The innovative finite element analysis module integration solution, through refined grid division and dynamic boundary condition setting, can accurately simulate the stress distribution of equipment under various working conditions, accurately locate high-stress areas and potential fault points, provide precise guidance for equipment maintenance, and significantly reduce equipment failure rate and maintenance costs; 4. The system adopts the Runge-Kutta numerical method and Kalman filter algorithm, has good adaptability and anti-interference ability, and can adapt to the complex and changeable operating environment of the power grid. Through continuous learning and model optimization, the system's early warning accuracy and robustness are continuously improved, providing reliable protection for the safe operation of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0047] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0048] This invention combines digital twin technology, finite element analysis and machine learning isolation forest algorithm to provide a comprehensive power grid equipment fault abnormality warning design solution. Figure 1 As shown, the overall process is:

[0049] By collecting and integrating multi-source data, a digital twin model is built and updated in real time, enabling dynamic monitoring and status simulation of power grid equipment. Furthermore, a finite element analysis module is used to analyze equipment behavior under various stress conditions in detail, simulating potential high-stress areas and failure points. The isolation forest model, using unsupervised learning methods, quickly and effectively detects abnormal data, providing timely warnings of equipment failures.

[0050] Step 1: Data collection and preprocessing

[0051] First, it is necessary to collect various operating data of the power grid system, including historical data and real-time data. These data can come from multiple sources, such as sensors, smart meters, SCADA (Supervisory Control and Data Acquisition) systems, etc. The collected data include parameters such as current, voltage, frequency, temperature, humidity, as well as equipment health status data (such as transformer oil temperature, winding temperature, insulation status, vibration, etc.), environmental data (such as meteorological data: wind speed, rainfall, temperature, humidity, and environmental pollution data: PM2.5, SO2 concentration, etc.) and load data (such as load current, load voltage, power factor, etc.). After data collection, data cleaning and preprocessing are performed to remove noise and outliers to ensure the accuracy and integrity of the data.

[0052] Step 2: Establishment of digital twin model

[0053] The core concept of digital twin technology is to replicate and simulate the actual operation of the power grid by establishing a digital model corresponding to the physical system of the power grid, so as to achieve a deep understanding of the system and real-time optimization. In the method of early identification and warning of power grid equipment defects and anomalies, the establishment of the digital twin model includes:

[0054] Modeling the various components of a power grid as mathematical equations captures the physical characteristics and behavior of the grid. The topology of a power grid can be represented as nodes and edges in graph theory, where nodes represent grid elements (such as generators, transformers, and distribution lines), and edges represent the connections between grid elements. The mathematical equations that establish these connections are as follows:

[0055] G=(N,E)

[0056] Among them, G is the power grid graph, N is the node set, and E is the edge set. The attributes of the nodes and edges are used to represent the resistance, reactance, load and other parameters of the components. State variable modeling uses the power flow equation to describe the voltage and power distribution of the power grid to simulate the current flow and energy transmission in the power grid, which is crucial for the operation and management of the power grid. Power grid simulation is the process of simulating the operating behavior of the power grid through a model, which includes initial condition setting, time stepping simulation and real-time data integration. The setting of initial conditions involves setting initial values ​​for the power grid model, including node voltage and power. Power grid simulation usually adopts the time stepping method, starting from the initial conditions, and gradually simulating the operation of the power grid over a period of time. The Runge-Kutta numerical method is used in the process to solve the power flow equation. The power flow equation is as follows:

[0057] P i -P j =V i ·V j ·(G ij ·cos(θ i -θ j )+B ij ·sin(θ i -θ j ))

[0058] Q i -Q j =V i ·V j ·(G ij ·sin(θ i -θ j )-B ij ·cos(θ i -θ j ))

[0059] Where, P i and Q i are the active and reactive power at node i, V i is the voltage at node i, θ i θ j is the phase angle between nodes i and j, G ij and B ij are the real and imaginary parts of the admittance matrix respectively.

[0060] However, the state and configuration of the power grid change over time, necessitating regular updates to the digital twin model. This involves data acquisition and processing, integrating large amounts of data collected from various sources. By comparing actual operating data with the model's predictions, discrepancies can be identified and the model can be updated. This process is accomplished using Kalman filtering technology, which combines actual measurements with model predictions to produce a more accurate model. The Kalman filter equation is shown below:

[0061] x k =A·x k -1+B·u k +w k

[0062] Where x k is the state vector, A can be expressed as node voltage and power in the power grid; B is the state transfer matrix; u k is the control input matrix; w k The Kalman filter models process noise by considering state transitions and control inputs, thereby better tracking the state changes of the power grid.

[0063] Step 3: Simulate abnormal equipment failure based on finite element analysis

[0064] By integrating the finite element analysis module into the digital twin model, abnormal and fault conditions of power grid equipment can be simulated.

[0065] (31) Establish finite element analysis module

[0066] Past fish operation data in the digital twin model can be used as input conditions for finite element analysis to define the boundary conditions and loading conditions of the equipment in actual operation.

[0067] Then, based on the geometric structure and material properties of the equipment, a detailed finite element module is integrated, and the finite element analysis software ANSYS is used for meshing to analyze the stress conditions that cause deformation or damage to the equipment under different conditions. Understanding the stress distribution helps predict possible failure points and high-risk areas of the equipment.

[0068] The geometry should include the shape, size, and specific geometric features of the device, such as the winding structure of the transformer, the connection points of the bracket, and the thickness of the shell. These geometric details determine the response of the device under stress. Material properties must cover the physical properties of each part of the device, such as elastic modulus, density, Poisson's ratio, and thermal conductivity. These parameters are critical for simulating the accurate behavior of the device under different stress conditions. Adding refined meshing to the existing digital twin model so that the finite element model can more accurately describe the structural characteristics of the device. The refined mesh helps capture smaller geometric features and local stress concentration areas, thereby improving the accuracy and reliability of the simulation.

[0069] (32) Setting boundary conditions and loading conditions

[0070] Based on the equipment's actual operating data, the various forces and stresses to which the equipment is subjected during operation, including mechanical loads, electromagnetic forces, and thermal stresses, are defined. Boundary conditions define the constraints on the equipment, such as fixed support points, free boundaries, and sliding supports, to restrict certain degrees of freedom. By accurately defining boundary and loading conditions, finite element analysis can simulate the stress distribution and deformation of the equipment under normal and abnormal operating conditions, thereby identifying potential failure points and high-stress areas. Finite element analysis software is used to solve and analyze the stress and deformation data of the equipment under various operating conditions.

[0071] Step 4: Equipment failure and abnormality warning based on isolation forest

[0072] Building upon the digital twin, machine learning models can be further integrated to provide early warnings of equipment failures. By isolating normal samples from abnormal samples without requiring prior labels, anomalous behavior in the data can be quickly and effectively identified. The Isolation Forest model can provide early warnings of failures without labeled data. For many equipment monitoring systems, acquiring large amounts of labeled data can be difficult or expensive. However, the Isolation Forest model avoids this problem by building a model based solely on normal operating data.

[0073] The Isolation Forest model has the ability to quickly detect anomalies. It uses a tree-based approach to construct an isolation tree by randomly selecting features and randomly splitting data, effectively identifying outliers. When using the established isolation forest to judge the anomaly of a sample, it is first necessary to determine the number of stopping levels of the sample in each binary tree in the isolation forest. The stopping level refers to the path length of the sample in each binary tree, that is, the number of levels from the root node to the leaf node where the sample is located. Generally speaking, the smaller the number of stopping levels, the more likely the sample is to be considered an outlier, as shown in the following formula:

[0074]

[0075] Where E|h(x)| represents the expected value of the path length (average path length) in the isolation forest, and the calculation formula of c(ψ) is as follows:

[0076]

[0077] Among them, ψ is a key parameter of the power grid equipment status, reflecting the equipment operating status; H(ψ-1) is the Heaviside step function, which is used to determine whether ψ is greater than 2; n is a normalization constant that affects the influence of ψ on the result; c(ψ) is used to describe the anomaly detection function in the power grid equipment fault early warning system.

[0078] It can be observed that the lower the anomaly score, the more isolated the data point, and thus the more likely it is an outlier. This is because the core concept of the Isolation Forest Model is to separate normal and abnormal points by constructing a tree-like structure. In this tree structure, abnormal points tend to have shorter path lengths because they are relatively far apart in the feature space, making them more likely to be "isolated" than most normal points. By setting a threshold, data points below this threshold are identified as possible abnormal failures, thus providing timely early warning of equipment failures. The early warning results are monitored and visualized in real time through the digital twin system. Operators can take appropriate maintenance measures based on the warning information to avoid further problems caused by equipment failures.

[0079] The invention also discloses an electronic device.

[0080] Specifically, the electronic device can be a computer device such as a desktop computer, a laptop computer, a PDA, and a cloud server. The computer device may include, but is not limited to, a processor and a memory. The processor and the memory may be connected via a bus or other means. The processor may be a central processing unit (CPU). The processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, graphics processing units (GPU), embedded neural network processors (NPU) or other dedicated deep learning coprocessors, discrete gate or transistor logic devices, discrete hardware components and other chips, or a combination of the above-mentioned chips.

[0081] As a non-transient computer-readable storage medium, the memory can be used to store non-transient software programs, non-transient computer executable programs and modules. The processor executes various functional applications and data processing of the processor by running the non-transient software programs, instructions and modules stored in the memory. The memory may include a program storage area and a data storage area, wherein the program storage area may store a control unit, an application required for at least one function; the data storage area may store data created by the processor, etc. In addition, the memory may include a high-speed random access memory and may also include a non-transient memory. In some embodiments, the memory may optionally include a memory remotely located relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network and a combination thereof.

[0082] The invention also discloses a computer-readable storage medium.

[0083] Specifically, a computer-readable storage medium is used to store a computer program, and when the computer program is executed by a processor, the method in the above-mentioned method implementation is implemented. Those skilled in the art will understand that the implementation of all or part of the process in the above-mentioned embodiment method of the present application can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. When the program is executed, it may include the process of the implementation of each of the above-mentioned methods. Among them, the storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory (Flash Memory), a hard disk (Hard Disk Drive, abbreviated: HDD) or a solid-state drive (SSD), etc.; the storage medium may also include a combination of the above-mentioned types of memories.

Claims

1. A method for early identification and warning of abnormal defects in power grid equipment, characterized in that: The following steps are involved: (1) Collect the operating data of power grid equipment, including historical data and real-time data, and perform data cleaning and preprocessing; (2) Based on the pre-processed operating data, a digital twin model of the power grid equipment is established, and the model is updated in real time using Kalman filtering technology; (3) Integrate the finite element analysis module into the digital twin model to simulate the stress distribution of the equipment under different operating conditions and identify potential failure points; (4) Anomaly detection is performed on equipment operation data based on the isolation forest algorithm, and an early warning signal is generated when an anomaly is detected.

2. The early identification and warning method according to claim 1, characterized in that: The operating data in step 1 includes current, voltage, frequency, temperature, humidity, equipment health status data, environmental data and load data.

3. The early identification and warning method according to claim 1, characterized in that: The establishment of the digital twin model in step 2 includes: The grid topology is modeled as a graph theory model G = (N, E), where N represents the set of grid element nodes, E represents the set of connecting edges between elements, and the attributes of nodes and edges include resistance, reactance, and load parameters; A power grid operation state model is established based on the power flow equation, which is expressed as: P i -P j =V i ·V j ·(G ij ·cos(θ i -θ j )+B ij ·sin(θ i -θ j )) Q i -Q j =V i ·V j ·(G ij ·sin(θ i -θ j )-B ij ·cos(θ i -θ j )) Where, P i and Q i are the active and reactive power at node i, V i is the voltage at node i, θ i θ j is the phase angle between nodes i and j, G ij and B ij are the real and imaginary parts of the admittance matrix respectively.

4. The early identification and warning method according to claim 3, characterized in that: The solution of the power flow equation adopts the Runge-Kutta numerical method to perform time stepping simulation. The simulation process includes setting the initial conditions of node voltage and power and simulating the dynamic operation process of the power grid through iterative calculation.

5. The early identification and warning method according to claim 3, characterized in that: The update of the digital twin model is implemented using the Kalman filter algorithm. The calculation equation of the Kalman filter is: x k =A·x k -1+B·u k +w k Where x k is the state vector, A is represented by node voltage and power in the power grid; B is the state transfer matrix; u k is the control input matrix; w k For control input.

6. The early identification and warning method according to claim 1, characterized in that: The establishment of the finite element analysis module described in step 3 includes: based on the equipment operation data output by the digital twin model, constructing a finite element model including the equipment geometry and material properties, using the finite element analysis software ANSYS for meshing, and analyzing the stress conditions that cause deformation or damage to the equipment under different conditions; the geometric structure includes shape, size and key characteristic parameters; the material properties include elastic modulus, Poisson's ratio and thermal conductivity; according to the actual operation data of the equipment, setting mechanical load, electromagnetic force and thermal stress as boundary conditions, calculating the stress distribution of the equipment under various working conditions through finite element analysis, and identifying potential failure points.

7. The early identification and warning method according to claim 1, characterized in that: The anomaly detection of the equipment operation data based on the isolation forest algorithm described in step 4 includes: An isolation forest model consisting of multiple isolated trees is constructed, and the anomaly score S(x) is obtained by calculating the path length h(x) of the sample in the tree structure. The calculation formula is: Where E|h(x)| is the average path length of sample x in the isolation forest, c(ψ) is the normalization factor, and its calculation formula is: Where ψ is a key parameter of the power grid equipment status, reflecting the equipment operating status; H(ψ-1) is the Heaviside step function, used to determine whether ψ is greater than 2; n is a normalization constant that affects the impact of ψ on the result; c(ψ) is used to describe the anomaly detection function in the power grid equipment fault early warning system. When the abnormality score S(x) is lower than the preset threshold, it is determined that the equipment is abnormal and an early warning signal is generated.

8. An early warning system for identifying and warning of power grid equipment defects and abnormalities, characterized by: include: Data acquisition and preprocessing module, used to collect operating data of power grid equipment, including historical data and real-time data, and perform data cleaning and preprocessing; The digital twin modeling module is used to build a digital twin model of power grid equipment based on pre-processed operating data and update the model in real time using Kalman filtering technology; Finite element analysis module, integrated into the digital twin model, is used to simulate the stress distribution of equipment under different operating conditions and identify potential failure points; The anomaly detection and early warning module is used to detect anomalies in equipment operation data based on the isolation forest algorithm and generate an early warning signal when an anomaly is detected.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for early identification and warning of power grid equipment defects and abnormalities according to any one of claims 1 to 7 is implemented.

10. An electronic device comprising a memory, a processor, and a program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method for early identification and warning of power grid equipment defects and abnormalities according to any one of claims 1 to 7 is implemented.

Images