Intelligent power grid equipment state evaluation and prediction method based on artificial intelligence

Abstract

The invention relates to the technical field of power grid equipment management, in particular to a power grid equipment state intelligent evaluation and prediction method based on artificial intelligence. Through data acquisition and processing and a real-time monitoring and fault early warning mechanism, in the data acquisition stage, historical data are called, key electrical parameters and fault information are acquired in real time by means of a sensor, a data and fault association mechanism is established, and in the real-time monitoring link, the similarity between a real-time curve and a standardized curve is calculated, so that the fault early warning accuracy is improved. Once the similarity reaches a threshold value, the system quickly gives an early warning, preliminarily judges the fault position, more accurately detects the specific condition of the fault position, and meanwhile, performs troubleshooting according to the weight of the fault position, can quickly position the fault and deal with the fault in time, reduces the power failure time of the fault, guarantees the stable power supply of a power grid, and reduces the influence on the production and life of a user. The method is of great significance in improving the safe operation level of the whole power grid.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid equipment management, and in particular to an artificial intelligence-based intelligent evaluation and prediction method for power grid equipment status. Background Art

[0002] With the rapid development of power systems and the continuous expansion of power grids, the stable operation of grid equipment is crucial to ensuring power supply. However, over the long term, grid equipment is susceptible to various factors, such as equipment aging, environmental changes, and load fluctuations, leading to various faults such as short circuits, grounding, and overloads. These faults not only affect the continuity and reliability of power supply, but can also cause equipment damage, increased repair costs, and even lead to safety accidents.

[0003] Traditional methods for assessing and predicting the condition of power grid equipment primarily rely on manual inspections and scheduled maintenance. Manual inspections are inefficient, susceptible to subjective factors, and difficult to detect potential problems. Scheduled maintenance, on the other hand, often lacks specificity and can lead to over- or under-maintenance. Furthermore, traditional methods are limited in processing large amounts of complex data, making it difficult to accurately analyze equipment operating status and failure trends. Therefore, developing an AI-based intelligent method for assessing and predicting the condition of power grid equipment is crucial for improving the operating efficiency of power grid equipment, reducing failure risks, and ensuring safe and stable grid operation. Summary of the Invention

[0004] To achieve the above objectives, the present invention proposes an artificial intelligence-based intelligent evaluation and prediction method for power grid equipment status, comprising the following steps:

[0005] Step 1: Data collection and correlation

[0006] Collect historical and real-time voltage and current data and corresponding fault information of power grid equipment, match the electrical data at the time of the fault through timestamps, establish the association between the fault location and electrical parameters based on the power grid topology, and generate electrical parameter curves before and after the fault;

[0007] Step 2: Construction of normalized curve

[0008] Normalize multiple sets of electrical parameter curves for the same type of faults, calculate the mean at the same time point to generate a standardized curve, set the threshold range based on the peak statistics and standard deviation of historical fault data, and store the standardized curve and threshold;

[0009] Step 3: Real-time fault warning and location

[0010] Generate electrical parameter curves in real time, calculate their similarity with the standardized curves through the dynamic time warping algorithm, and trigger an early warning when the similarity is greater than or equal to the set threshold;

[0011] Use support vector machines to analyze electrical data to initially locate faults, combine camera images with convolutional neural networks to verify the fault location, and optimize the troubleshooting sequence based on equipment fault frequency weights.

[0012] Step 4: Failure risk prediction

[0013] Based on the Markov prediction model, the equipment status is defined and the state transition matrix is ​​constructed. The probability of the equipment transitioning to a faulty state in the future is calculated based on the current state. When the probability exceeds the risk threshold, maintenance is arranged in advance.

[0014] Step 5: Post-repair status assessment

[0015] The electrical parameters after maintenance are collected to generate an operating curve. The features are extracted through a recurrent neural network combined with Fourier transform and compared with the standard normal curve. If the deviation exceeds the threshold, secondary maintenance is initiated.

[0016] In one example, the normalization process in step 2 adopts the Z-Score normalization method, and the specific formula is:

[0017] Among them, is the original data, is the mean of the data set, is the standard deviation of the data set, and is the normalized data.

[0018] In one example, the step of generating the normalization curve includes the following steps:

[0019] S2.1. Collect multiple sets of voltage and current data under the same type of fault;

[0020] S2.2. Normalize each set of data;

[0021] S2.3. Calculate the mean and standard deviation of each set of data at the same time point;

[0022] S2.4. Generate a standardized curve based on the average value;

[0023] S2.5. Determine the fluctuation range of the curve based on the standard deviation and form a curve range diagram.

[0024] In one example, the threshold range in step 2 is determined by statistically analyzing the minimum peak and maximum peak values ​​of the current in similar fault cases, calculating the mean and standard deviation of historical fault data, setting a reasonable multiple of the standard deviation range as the threshold range, and optimizing and adjusting the threshold in combination with the operating experience of the equipment and expert knowledge.

[0025] In one example, the weight determination in step three includes the following steps:

[0026] S3.1. Collect historical fault data and count the frequency of faults at each equipment location;

[0027] S3.2. Calculate the fault frequency weight of each device location using the following formula: Among them, W i is the failure frequency weight of the equipment location, i is the frequency of failure at the equipment location, and n is the total number of equipment locations;

[0028] S3.3. Combining expert knowledge and equipment importance, adjust the fault frequency weight to obtain the final fault location weight.

[0029] In one example, the steps of the Markov prediction model in step 4 are as follows:

[0030] S4.1. Define the operating status of the equipment, which usually includes normal status, warning status, and fault status;

[0031] S4.2. Collect historical operation data and count the transition frequencies between different states;

[0032] S4.3. Construct the state transfer matrix P, where P i,j represents the probability of transitioning from state i to state j;

[0033] S4.4. Predict the future state of the device based on the current state and the state transition matrix.

[0034] In one example, the current state in the fault risk prediction is S current , using the state transition matrix P, calculate the probability P of the device transferring to the fault state in the future time t fault (t), set the risk threshold R threshold , when P fault (t)≥R threshold When the equipment is at risk of failure, it is determined that there is a risk of failure, an early warning is issued and maintenance is arranged in advance.

[0035] In one example, the specific process of the post-repair status assessment in step 5 includes the following steps:

[0036] S5.1. Collect the voltage and current data of the equipment again and generate the normal operation curve;

[0037] S5.2. Use time series data analysis algorithms such as RNN or LSTM, combined with Fourier transform signal processing technology to extract features;

[0038] S5.3. Compare and determine the extracted features with the standard normal curve features before maintenance;

[0039] S5.4. If the deviation between the curve and the standard curve exceeds the set threshold range, it indicates that the maintenance may not be thorough enough and further inspection of various areas of the equipment is required;

[0040] S5.5. Combine the equipment maintenance records and historical operation data to conduct a comprehensive analysis of the test results to identify the specific causes of the curve anomaly;

[0041] S5.6. Carry out key inspections on abnormal areas and improve maintenance work until the equipment's operating curve returns to normal.

[0042] The artificial intelligence-based intelligent evaluation and prediction method for power grid equipment status proposed by the present invention can bring the following beneficial effects:

[0043] 1. The present invention can effectively improve the reliability and safety of power grid equipment operation through data collection and processing, as well as real-time monitoring and fault warning mechanisms. In the data collection stage, not only historical data is retrieved, but also key electrical parameters and fault information are obtained in real time with the help of sensors, and a data-fault correlation mechanism is established, laying a solid foundation for subsequent analysis and modeling. In the real-time monitoring link, the DTW algorithm is used to calculate the similarity between the real-time curve and the standardized curve. Once the similarity reaches the threshold, the system quickly issues an early warning and combines SVM to preliminarily determine the fault location. It also combines on-site image data with algorithms such as CNN to more accurately detect the specific conditions of the fault location. At the same time, it prioritizes the inspection based on the weight of the fault location, can quickly locate the fault and handle it in a timely manner, reduce the power outage time, ensure the stable power supply of the power grid, reduce the impact on users' production and life, and is of great significance to improving the safe operation level of the entire power grid.

[0044] 2. In terms of fault risk prediction, the present invention collects and standardizes pre-fault data, analyzes regular characteristics to set threshold ranges, continuously compares the operating curve with the pre-fault standardized curve, and uses the Markov prediction model to predict fault risks based on similarity and threshold ranges, arranges maintenance in advance, takes preventive measures, rationally plans maintenance resources, and avoids large-scale maintenance and economic losses caused by sudden equipment failures. During post-maintenance inspection and evaluation, data is collected again to generate a normal operating curve, which is compared with the standard normal curve before maintenance. RNN or LSTM is combined with Fourier transform to extract features to accurately determine whether the equipment has resumed normal and stable operation. If the deviation is large, further maintenance is required to ensure maintenance quality, extend the service life of the equipment, improve the overall operating efficiency and economic benefits of the power grid equipment, and provide strong technical support for the long-term stable operation of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0046] Figure 1Schematic diagram of the flow chart of this AI-based intelligent assessment and prediction method for power grid equipment status. DETAILED DESCRIPTION

[0047] In order to more clearly illustrate the overall concept of the present invention, a detailed description is given below in an exemplary manner in conjunction with the accompanying drawings.

[0048] In the description of the present invention, it should be understood that the terms "center", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "axial", "radial", "circumferential", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the present invention.

[0049] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature identified as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0050] In the present invention, unless otherwise expressly specified or limited, terms such as "mounted," "connected," "connect," and "fixed" should be understood broadly. For example, they may refer to fixed connection, detachable connection, or integration; mechanical connection, electrical connection, or communication; direct connection or indirect connection through an intermediate medium; and internal communication between two components or interaction between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0051] In the present invention, unless otherwise clearly specified and limited, a first feature "above" or "below" a second feature may be that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. In the description of this specification, the descriptions with reference to the terms "one scheme", "some schemes", "examples", "specific examples", or "some examples" mean that the specific features, structures, materials or characteristics described in conjunction with the scheme or example are included in at least one scheme or example of the present invention. In this specification, the schematic expressions of the above terms do not necessarily refer to the same scheme or example. Moreover, the specific features, structures, materials or characteristics described may be combined in an appropriate manner in any one or more schemes or examples.

[0052] like Figure 1As shown, the present invention proposes an artificial intelligence-based intelligent evaluation and prediction method for power grid equipment status, comprising the following steps:

[0053] Step 1: Data collection and preliminary processing. During the data collection process, historical power grid data is first retrieved, including the voltage, current and other electrical parameter data of the equipment, as well as related fault information when the equipment fails. Through various sensors installed on the power grid equipment, the voltage, current and other electrical parameter data of the equipment, as well as related fault information when the equipment fails, such as fault type and fault occurrence time, are obtained in real time. It is key to match the fault information with the voltage and current data.

[0054] During data processing, a mechanism for associating data with faults is established to ensure that the voltage and current data at the corresponding moment can be accurately found when a fault occurs. Timestamps can be used for matching, and combined with the topological structure information of the power grid, the fault location and related electrical data can be more accurately associated. These data can be plotted as curves, and data from specific time periods before and after the fault can be captured for plotting, providing a basis for subsequent analysis and modeling.

[0055] Step 2: Data normalization and standardized curve generation: normalize multiple sets of voltage and current curves under the same fault type to eliminate the influence of different data dimensions and units.

[0056] Data normalization is to unify data of different dimensions and dimensional units so that they fall into a specific range. Z-Score normalization is used to convert data into a standard normal distribution by calculating the mean and standard deviation of the data. The formula is as follows:

[0057] Among them, X is the original data, μ is the mean of the data set, σ is the standard deviation of the data set, X norm It is normalized data. Normalization can solve the problem of large data differences between different devices and under different operating conditions, making the data comparable on a unified scale.

[0058] Based on the normalized data, a standardized curve is generated, which represents the typical pattern of voltage and current changes of the equipment under this type of fault.

[0059] The steps to generate a normalization curve are as follows:

[0060] S2.1. Collect multiple sets of voltage and current data under the same type of fault;

[0061] S2.2. Normalize each set of data;

[0062] S2.3. Calculate the mean and standard deviation of each set of data at the same time point;

[0063] S2.4. Generate a standardized curve based on the average value;

[0064] S2.5. Determine the fluctuation range of the curve based on the standard deviation and form a curve range diagram.

[0065] Statistically analyze the minimum and maximum peak values ​​of the current in similar fault cases to determine the preliminary threshold range. Statistically analyze the minimum and maximum peak values ​​of the current in similar fault cases, calculate the mean and standard deviation of historical fault data, and set a reasonable multiple of the standard deviation range as the threshold range, usually taking the mean ± 2 times the standard deviation. Combined with equipment operating experience and expert knowledge, optimize and adjust the threshold, store the standardized curve and related threshold range information in the database to provide a reference for real-time monitoring and fault warning.

[0066] Step 3: Real-time monitoring and fault warning: Collect current and voltage data from power grid equipment in real time and generate real-time curves. Use the Dynamic Time Warping (DTW) algorithm to calculate the similarity between the real-time curves and the standardized curves in the database.

[0067] When the similarity reaches the threshold, the system issues an early warning. Combined with the fault information, the real-time curve is set as X and the standardized curve is set as Y. The DTW algorithm is used to calculate the cumulative distance DTW between X and Y. SCORE , the cumulative distance DTW SCORE Converted to similarity S, the conversion formula is as follows

[0068] The similarity S ranges from (0, 1). The closer the value is to 1, the higher the similarity is. The similarity threshold S is set. threshold , when the real-time curve S and the standardized curve S threshold When the similarity is greater than or equal to , the system issues an early warning.

[0069] Use artificial intelligence algorithms such as support vector machines (SVM) to preliminarily determine the fault location. Combine image data collected by on-site cameras with image recognition algorithms such as convolutional neural networks (CNN) to detect the specific condition of the fault location. Prioritize fault locations based on their weights, improving troubleshooting efficiency.

[0070] The steps for weight determination are as follows:

[0071] S3.1. Collect historical fault data and count the frequency of faults at each equipment location;

[0072] S3.2. Calculate the fault frequency weight of each device location using the following formula: Among them, W i is the failure frequency weight of the equipment location, i is the frequency of failure at the equipment location, and n is the total number of equipment locations;

[0073] S3.3. Combining expert knowledge and equipment importance, adjust the fault frequency weight to obtain the final fault location weight.

[0074] The threshold determination can be dynamically adjusted based on the real-time operating status of the equipment and environmental factors to ensure the accuracy and timeliness of the early warning.

[0075] Step 4: Fault risk prediction: Collect and standardize the voltage and current data before the equipment fails. Analyze its regular characteristics and set the threshold range. Continuously collect the current and voltage data during operation to generate a curve, which is compared with the standardized curve before the failure. Using algorithms such as the Markov prediction model, based on the similarity between the current operating curve and the pre-fault curve and the threshold range, predict the fault risk and arrange maintenance in advance. The steps of the Markov prediction model are as follows:

[0076] S4.1. Define the operating status of the equipment, which usually includes normal status, warning status, and fault status;

[0077] S4.2. Collect historical operation data and count the transition frequencies between different states;

[0078] S4.3. Construct the state transfer matrix P, where P i,j represents the probability of transitioning from state i to state j;

[0079] S4.4. Predict the future state of the device based on the current state and the state transition matrix.

[0080] Using the Markov prediction model, combined with the current operating status of the equipment and the state transition matrix, the equipment failure risk is assessed. The failure risk assessment process is as follows:

[0081] Let the current state be S current , using the state transition matrix P, calculate the probability P of the device transferring to the fault state in the future time t fault (t), set the risk threshold R threshold , when P fault (t)≥R threshold When the equipment is at risk of failure, it is determined that there is a risk of failure, an early warning is issued and maintenance is arranged in advance.

[0082] Step 5: Post-repair inspection and evaluation. After repair, the equipment's voltage and current data are collected again to generate a normal operating curve. This is then compared with the standard normal curve before repair. Deep learning algorithms such as recurrent neural networks (RNNs) or long short-term memory networks (LSTMs) are used to analyze time series data, combined with signal processing techniques such as Fourier transforms to extract features and accurately determine whether the curve meets the standard. If the deviation is significant, further inspection and evaluation are required to ensure normal and stable operation of the equipment after repair.

[0083] Recurrent Neural Network (RNN) is a type of neural network used to process sequence data and can capture dynamic information in time series. However, traditional RNN has the problem of gradient vanishing and gradient exploding. Long Short-Term Memory Network (LSTM) effectively solves this problem by introducing a gating mechanism. The structure of LSTM includes input gate, forget gate and output gate, which can selectively remember or forget information, thereby better processing long sequence data. Combined with signal processing techniques such as Fourier transform to extract features, it can more accurately reflect the actual operating status of the device. Fourier transform is a mathematical tool that converts time domain signals into frequency domain signals, which can reveal the periodic components in the signal. The formula of Fourier transform is as follows

[0084] Among them, x(t) is the time domain signal, X(f) is the frequency domain signal, and f is the frequency. Through Fourier transform, the voltage and current signals can be converted from the time domain to the frequency domain, and the main frequency components and amplitudes in the signals can be analyzed to extract features for equipment status assessment.

[0085] Post-repair inspection and evaluation process:

[0086] S5.1. Collect the voltage and current data of the equipment again and generate the normal operation curve;

[0087] S5.2. Use time series data analysis algorithms such as RNN or LSTM, combined with signal processing techniques such as Fourier transform to extract features.

[0088] S5.3. Compare and judge the extracted features with the standard normal curve features before maintenance.

[0089] S5.4. If the deviation between the curve and the standard curve exceeds the set threshold range, it indicates that the maintenance may not be thorough enough and further inspection of various areas of the equipment is required.

[0090] S5.5. Combine the equipment maintenance records and historical operating data to conduct a comprehensive analysis of the test results to find out the specific reasons for the abnormal curve.

[0091] S5.6. Carry out key inspections on abnormal areas and improve maintenance work until the equipment's operating curve returns to normal.

[0092] This system achieves intelligent management of the entire process through a five-stage closed-loop approach. First, a dynamic correlation mechanism for multi-source data is established to accurately link electrical parameters and fault characteristics at the time of a fault, generating a normalized standard curve library. During the real-time monitoring phase, a dynamic time warping algorithm is used to compare the real-time curve with the standard curve. Graded warnings are triggered using two-level similarity thresholds of 85% and 92%. A dual-modal fault location model (70% electrical weighting and 30% visual weighting) is formed by combining electrical feature analysis using a support vector machine with visual recognition using a convolutional neural network. This model accurately locates the fault point and generates troubleshooting priorities. A Markov model is used to quantify equipment state transition probabilities, predicting failure risks within 40 minutes and scheduling maintenance in advance. Finally, an LSTM network is used to integrate frequency domain features for in-depth evaluation of maintenance effectiveness, triggering secondary maintenance for abnormal equipment. The diagnostic results are then fed back to the standard curve library and algorithm model for system self-optimization. This approach addresses core pain points in traditional operation and maintenance, such as delayed warnings, ambiguous positioning, and inadequate assessments, significantly improving the timeliness, accuracy, and reliability of power grid equipment state management.

[0093] The various embodiments in this specification are described in a progressive manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the system embodiments are generally similar to the method embodiments, so the description is relatively simple. For relevant parts, refer to the description of the method embodiments.

[0094] The foregoing is merely an embodiment of the present invention and is not intended to limit the present invention. It will be apparent to those skilled in the art that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention are intended to be included within the scope of the claims of the present invention.

Claims

1. An artificial intelligence-based intelligent assessment and prediction method for power grid equipment status, characterized by: The following steps are involved: Step 1: Data collection and correlation Collect historical and real-time voltage and current data and corresponding fault information of power grid equipment, match the electrical data at the time of the fault through timestamps, establish the association between the fault location and electrical parameters based on the power grid topology, and generate electrical parameter curves before and after the fault; Step 2: Normalization curve construction Normalize multiple sets of electrical parameter curves for the same type of faults, calculate the mean at the same time point to generate a standardized curve, set the threshold range based on the peak statistics and standard deviation of historical fault data, and store the standardized curve and threshold; Step 3: Real-time fault warning and location Generate electrical parameter curves in real time, calculate their similarity with the standardized curves through the dynamic time warping algorithm, and trigger an early warning when the similarity is greater than or equal to the set threshold; Use support vector machines to analyze electrical data to initially locate faults, combine camera images with convolutional neural networks to verify the fault location, and optimize the troubleshooting sequence based on equipment fault frequency weights. Step 4: Failure risk prediction Based on the Markov prediction model, the equipment status is defined and the state transition matrix is ​​constructed. The probability of the equipment transitioning to a faulty state in the future is calculated based on the current state. When the probability exceeds the risk threshold, maintenance is arranged in advance. Step 5: Post-repair status assessment The electrical parameters after maintenance are collected to generate an operating curve. The features are extracted through a recurrent neural network combined with Fourier transform and compared with the standard normal curve. If the deviation exceeds the threshold, secondary maintenance is initiated.

2. The method for intelligent evaluation and prediction of power grid equipment status based on artificial intelligence according to claim 1 is characterized by: The normalization process in step 2 adopts the Z-Score normalization method, and the specific formula is: Among them, is the original data, is the mean of the data set, is the standard deviation of the data set, and is the normalized data.

3. The method for intelligent evaluation and prediction of power grid equipment status based on artificial intelligence according to claim 1 is characterized in that: The step of generating the standardized curve comprises the following steps: S2.

1. Collect multiple sets of voltage and current data under the same type of fault; S2.

2. Normalize each set of data; S2.

3. Calculate the mean and standard deviation of each set of data at the same time point; S2.

4. Generate a standardized curve based on the average value; S2.

5. Determine the fluctuation range of the curve based on the standard deviation and form a curve range diagram.

4. The method for intelligent evaluation and prediction of power grid equipment status based on artificial intelligence according to claim 1 is characterized in that: The threshold range in step 2 is determined by statistically analyzing the minimum and maximum peak values ​​of the current in similar fault cases, calculating the mean and standard deviation of historical fault data, setting a reasonable multiple of the standard deviation range as the threshold range, and optimizing and adjusting the threshold based on the equipment operating experience and expert knowledge.

5. The method for intelligent evaluation and prediction of power grid equipment status based on artificial intelligence according to claim 1 is characterized in that: The weight determination in step 3 includes the following steps: S3.

1. Collect historical fault data and count the frequency of faults at each equipment location; S3.

2. Calculate the fault frequency weight of each device location using the following formula: Among them, W i is the failure frequency weight of the equipment location, i is the frequency of failure at the equipment location, and n is the total number of equipment locations; S3.

3. Combining expert knowledge and equipment importance, adjust the fault frequency weight to obtain the final fault location weight.

6. The method for intelligent evaluation and prediction of power grid equipment status based on artificial intelligence according to claim 1, characterized in that: The steps of the Markov prediction model in step 4 are as follows: S4.

1. Define the operating status of the equipment, which usually includes normal status, warning status, and fault status; S4.

2. Collect historical operation data and count the transition frequencies between different states; S4.

3. Construct the state transfer matrix P, where P i,j represents the probability of transitioning from state i to state j; S4.

4. Predict the future state of the device based on the current state and the state transition matrix.

7. The method for intelligent evaluation and prediction of power grid equipment status based on artificial intelligence according to claim 1, characterized in that: In the fault risk prediction, the current state is assumed to be S current , using the state transition matrix P, calculate the probability P of the device transferring to the fault state in the future time t fault (t), set the risk threshold R threshold , when P fault (t)≥R threshold When the equipment is at risk of failure, it is determined that there is a risk of failure, an early warning is issued and maintenance is arranged in advance.

8. The method for intelligent evaluation and prediction of power grid equipment status based on artificial intelligence according to claim 1, characterized in that: The specific process of post-repair condition assessment in step 5 includes the following steps: S5.

1. Collect the voltage and current data of the equipment again and generate the normal operation curve; S5.

2. Use time series data analysis algorithms combined with Fourier transform signal processing technology to extract features; S5.

3. Compare and determine the extracted features with the standard normal curve features before maintenance; S5.

4. If the deviation between the curve and the standard curve exceeds the set threshold range, it indicates that the maintenance may not be thorough enough and further inspection of various areas of the equipment is required; S5.

5. Combine the equipment maintenance records and historical operation data to conduct a comprehensive analysis of the test results to identify the specific causes of the curve anomaly; S5.

6. Carry out key inspections on abnormal areas and improve maintenance work until the equipment's operating curve returns to normal.

Images