Power grid geological disaster meteorological risk prediction model construction and medium-term prediction method

Abstract

The invention relates to the technical field of power system disaster prevention and reduction, in particular to a power grid geological disaster meteorological risk prediction model construction and medium-term prediction method. According to the method, tower foundation deformation data and power grid operation data are innovatively introduced into the prediction model, and special tower branches and power grid branches are designed for processing; the fundamental transformation of power grid geological disaster risk prediction from only concerning an external meteorological geological environment to comprehensive consideration of external disaster-inducing factors and the state of a disaster-bearing body of a facility is realized. According to the multi-source data deep fusion model, the real risk of a specific tower under the severe weather condition can be evaluated more accurately, so that the prediction precision is remarkably improved, false alarm and missing alarm caused by neglecting the health condition of a facility individual in a traditional method are effectively reduced, and the prediction accuracy is improved. And a more reliable technical support is provided for a power grid to carry out targeted medium-term disaster prevention scheduling and operation and maintenance decisions.

Description

Technical Field

[0001] This invention relates to the field of disaster prevention and mitigation technology in power systems, specifically to a method for constructing a meteorological risk prediction model for geological disasters in power grids and for medium-term prediction. Background Technology

[0002] With the increasing frequency of global climate change and extreme weather events, geological disasters such as landslides, mudslides, and ground subsidence pose a growing threat to power grid facilities such as transmission lines and substations. Geological disasters not only cause direct losses such as transmission line interruptions and tower collapses, but can also trigger large-scale power outages, affecting the safe and stable operation of the power system. Traditional power grid geological disaster prediction mainly relies on short-term weather forecasts and geological surveys, and is mostly based on single data sources or simple statistical models. These methods are unable to cope with complex meteorological and geological conditions, and suffer from poor timeliness and low accuracy, failing to meet the actual needs of power grid disaster prevention and mitigation.

[0003] In recent years, with the continuous development of meteorological forecasting technology and the widespread application of machine learning algorithms, new ideas and methods have been provided for predicting geological disaster risks in power grids. Prediction models based on multi-source data fusion have gradually become a research hotspot. However, existing research mainly focuses on short-term predictions (1-3 days ahead) and lacks the ability to predict medium-term (7 days ahead) geological disaster meteorological risks applicable to power grids. Therefore, proposing a medium-term prediction method for geological disaster meteorological risks in power grids has significant research and application value. Summary of the Invention

[0004] This invention provides a model for predicting meteorological risks of geological disasters in power grids and a method for medium-term prediction, in order to solve the problem that existing technologies lack the ability to predict the medium-term (next 7 days) meteorological risks of geological disasters in power grids.

[0005] In a first aspect, the present invention provides a method for constructing a meteorological risk prediction model for geological disasters in power grids. The method includes: acquiring historical meteorological data, geological data, tower foundation deformation data, power grid operation data, and corresponding historical disaster data; extracting feature data from the meteorological data, geological data, tower foundation deformation data, and power grid operation data; and training a time-series network model containing tower branches and power grid branches using the feature data to obtain a meteorological risk prediction model for geological disasters in power grids. The tower branches are used to process the feature data of the tower foundation deformation data to obtain processed tower features, and the power grid branches are used to process the feature data of the power grid operation data to obtain processed power grid features. The time-series network model is used to process the feature data of the meteorological data and geological data, as well as the tower features and the power grid features, to output a risk value for a preset number of days in the future.

[0006] This invention innovatively incorporates tower foundation deformation data and power grid operation data into the prediction model, and designs dedicated tower and power grid branches for processing. This achieves a fundamental shift in power grid geological disaster risk prediction, moving beyond solely focusing on external meteorological and geological environments to comprehensively considering both external disaster-causing factors and the condition of the facilities themselves. This multi-source data deep fusion model can more accurately assess the true risk of specific towers under severe weather conditions, thereby significantly improving prediction accuracy and effectively reducing false alarms and missed alarms caused by neglecting the individual health status of facilities in traditional methods. This provides more reliable technical support for the power grid to conduct targeted medium-term disaster prevention scheduling and operation and maintenance decisions.

[0007] In one optional implementation, the power grid operation data includes line load, insulator leakage current, and tower grounding resistance. Extracting feature data from the meteorological data, geological data, tower foundation deformation data, and power grid operation data includes: extracting feature data from the meteorological data and geological data using a preset algorithm; determining the cumulative deformation of the tower based on the tower foundation deformation data to obtain feature data of the tower foundation deformation data; determining the power grid load mutation coefficient based on the line load, and determining the equipment degradation index based on the insulator leakage current and tower grounding resistance to obtain feature data of the power grid operation data.

[0008] This invention extracts specific features from power grid operation data (such as line load, insulator leakage current, and tower grounding resistance) to calculate quantitative indicators such as the power grid load mutation coefficient and equipment degradation index, thereby achieving a precise characterization of the power grid's own operating status and health. These features are deeply integrated with meteorological, geological, and tower deformation characteristics, enabling the risk assessment model to not only focus on external environmental disaster-causing factors but also reflect the real-time pressure and potential vulnerabilities within power grid facilities. This significantly improves the pertinence and accuracy of predicting meteorological risks of geological disasters to the power grid, providing a more comprehensive and reliable basis for operation and maintenance decisions.

[0009] In one optional implementation, the cumulative deformation of the tower is determined using the following formula:

[0010] In the formula, This represents the tower foundation deformation data at time t. This represents the tower foundation deformation data at time t-t0; The load mutation coefficient is expressed by the following formula:

[0011] In the formula, This represents the maximum load at time t. This represents the average load at time t; The equipment degradation index is determined using the following formula:

[0012] In the formula, This represents the insulator leakage current at time t. This indicates the maximum leakage current of the insulator. This represents the tower grounding resistance at time t. This indicates the maximum value of the tower's grounding resistance. and This indicates the preset weight.

[0013] This invention defines three core quantitative indicators: "cumulative tower deformation," "load mutation coefficient," and "equipment degradation index." These indicators transform previously scattered and heterogeneous power grid monitoring data (such as foundation deformation, load curves, and insulator condition) into standardized and calculable risk characteristics. These characteristic formulas accurately characterize changes in the structural stability of power facilities, sudden changes in operating stress, and the decline in equipment health. This provides clear and unified input for subsequent risk prediction models, significantly improving the early identification capability and objectivity of power grid geological hazard assessment.

[0014] In one optional implementation, extracting feature data from the meteorological data, geological data, tower foundation deformation data, and power grid operation data further includes: performing exponential weighted smoothing on the cumulative deformation of the tower to obtain the processed cumulative deformation; and using principal component analysis to filter the feature data to obtain the filtered features.

[0015] In this invention, by applying exponential weighted smoothing to the cumulative deformation of the towers, random noise and short-term fluctuations in the monitoring data can be effectively filtered out, thus revealing the long-term trend of tower deformation more clearly and accurately, providing more reliable feature inputs for structural stability assessment. Simultaneously, principal component analysis is used to filter multi-source feature data, significantly eliminating redundancy between features and reducing data dimensionality while retaining most of the key information. This greatly improves the analytical efficiency and generalization ability of subsequent models, making the entire power grid geological disaster risk prediction system more accurate and efficient.

[0016] In one alternative implementation, the loss function used during training is expressed by the following formula:

[0017] In the formula, Indicates the mean square error loss. This indicates the loss due to tower deformation constraint. Indicates load constraint loss. and This indicates the balancing weight.

[0018] In this invention, prior knowledge of power grid safe operation (such as tower deformation safety thresholds and load mutation thresholds) is embedded into the model training process in the form of mathematical constraints through this loss function. Based on the traditional mean square error loss, this function adds specific penalty terms for tower deformation exceeding limits and load mutation exceeding limits, and balances these penalties using adjustable weights. This forces the model to simultaneously consider the accuracy of predicted values ​​and the physical safety rules of the power grid when learning prediction patterns, thereby significantly improving the engineering rationality and reliability of the prediction results, effectively avoiding misjudgments of model outputs that violate safety regulations, and reducing the risk of power grid safe operation from an algorithmic perspective.

[0019] In one alternative implementation, the tower branch is used to assign weights to the cumulative deformation of the tower, and the grid branch is used to correlate the synergistic risks of the grid load mutation coefficient and the equipment degradation index.

[0020] In this invention, by using a weight allocation mechanism for pole branches and a collaborative risk assessment of power grid branches, the core risk dimensions of pole structural status and power grid operation status are focused on respectively. This not only achieves precise weighting of the risk of cumulative pole deformation, but also links the synergistic effect of load mutation and equipment degradation, thereby constructing a comprehensive risk perception model that simultaneously covers facility structure and system operation. This enhances the comprehensiveness and interpretability of power grid geological disaster risk early warning, and can provide more refined decision-making basis for the safe operation and maintenance of power systems.

[0021] Secondly, the present invention provides a method for medium-term prediction of meteorological risk of geological disasters in power grids. The method includes: acquiring meteorological data, geological data, tower foundation deformation data, and power grid operation data; extracting feature data from the meteorological data, geological data, tower foundation deformation data, and power grid operation data; and inputting the feature data into a meteorological risk prediction model for geological disasters in power grids constructed by the method for constructing a meteorological risk prediction model for geological disasters in power grids as described in the first aspect and any one of the first aspects of the present invention, to obtain risk values ​​for a preset number of days in the future.

[0022] This invention not only extends the early warning period to several days in the future, gaining valuable time for disaster prevention deployment, but more importantly, by introducing operational and deformation data that reflect the health status of the power grid facilities themselves, the risk assessment results are more accurate and realistic, significantly reducing false alarms and missed alarms caused by relying solely on meteorological and geological information, and providing scientific and forward-looking decision support for ensuring the safe and stable operation of the power grid.

[0023] In one optional implementation, the method further includes: determining a risk level based on the risk value of the preset number of future days; and generating a handling result for the power grid and towers based on the risk level.

[0024] This invention achieves a direct transformation from risk assessment results to specific response decisions. By automatically mapping the future risk values ​​predicted by the model to explicit "risk levels," and intelligently generating corresponding power grid dispatching and tower maintenance recommendations for different levels, this solution upgrades predictive warnings into actionable guidelines. This significantly enhances the initiative, standardization, and timeliness of risk response, enabling maintenance personnel to allocate resources and take measures such as reinforcement, inspection, or load adjustment in advance and accurately based on scientific classification conclusions. This effectively improves the precision of power grid disaster prevention and preparedness, ensuring the safe and stable operation of power facilities.

[0025] Thirdly, the present invention provides a device for constructing a meteorological risk prediction model for power grid geological disasters. The method includes: a historical data acquisition module for acquiring historical meteorological data, geological data, tower foundation deformation data, power grid operation data, and corresponding historical disaster data; a historical feature extraction module for extracting feature data from the meteorological data, geological data, tower foundation deformation data, and power grid operation data; and a model training and construction module for training a time-series network model containing tower branches and power grid branches using the feature data to obtain a meteorological risk prediction model for power grid geological disasters. The tower branches are used to process the feature data of the tower foundation deformation data to obtain processed tower features, and the power grid branches are used to process the feature data of the power grid operation data to obtain processed power grid features. The time-series network model is used to process the feature data of the meteorological data and geological data, as well as the tower features and the power grid features, to output a risk value for a preset number of days in the future.

[0026] In one optional implementation, the power grid operation data includes line load, insulator leakage current, and tower grounding resistance; the historical feature extraction module is specifically used to extract feature data from the meteorological and geological data using a preset algorithm; the cumulative deformation of the tower is determined based on the tower foundation deformation data to obtain feature data of the tower foundation deformation data; the power grid load mutation coefficient is determined based on the line load, and the equipment degradation index is determined based on the insulator leakage current and tower grounding resistance to obtain feature data of the power grid operation data.

[0027] In one optional implementation, the cumulative deformation of the tower is determined using the following formula:

[0028] In the formula, This represents the tower foundation deformation data at time t. This represents the tower foundation deformation data at time t-t0; The load mutation coefficient is expressed by the following formula:

[0029] In the formula, This represents the maximum load at time t. This represents the average load at time t; The equipment degradation index is determined using the following formula:

[0030] In the formula, This represents the insulator leakage current at time t. This indicates the maximum leakage current of the insulator. This represents the tower grounding resistance at time t. This indicates the maximum value of the tower's grounding resistance. and This indicates the preset weight.

[0031] In one optional implementation, the historical feature extraction module is specifically used to perform exponential weighted smoothing on the cumulative deformation of the tower to obtain the processed cumulative deformation; and to use principal component analysis to filter the feature data to obtain the filtered features.

[0032] In one alternative implementation, the loss function used during training is expressed by the following formula:

[0033] In the formula, Indicates the mean square error loss. This indicates the loss due to tower deformation constraint. Indicates load constraint loss. and This indicates the balancing weight.

[0034] In one alternative implementation, the tower branch is used to assign weights to the cumulative deformation of the tower, and the grid branch is used to correlate the synergistic risks of the grid load mutation coefficient and the equipment degradation index.

[0035] Fourthly, the present invention provides a medium-term prediction device for meteorological risks of geological disasters in power grids. The device includes: a data acquisition module for acquiring meteorological data, geological data, tower foundation deformation data, and power grid operation data; a feature extraction module for extracting feature data from the meteorological data, geological data, tower foundation deformation data, and power grid operation data; and a prediction module for inputting the feature data into a meteorological risk prediction model for geological disasters in power grids constructed by the method for constructing a meteorological risk prediction model for geological disasters in power grids according to the first aspect and any one of the first aspects of the present invention, to obtain risk values ​​for a preset number of days in the future.

[0036] In one optional embodiment, the device further includes: a risk level determination module, configured to determine a risk level based on the risk value of the preset number of future days; and a handling module, configured to generate handling results for the power grid and towers based on the risk level.

[0037] Fifthly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the above-described method for constructing a meteorological risk prediction model for power grid geological disasters or the method for medium-term prediction of meteorological risk of power grid geological disasters in the first aspect or any corresponding embodiment thereof, or the method for medium-term prediction of meteorological risk of power grid geological disasters in the second aspect or any corresponding embodiment thereof.

[0038] In a sixth aspect, the present invention provides a computer-readable storage medium storing computer instructions, which are used to cause a computer to execute the method for constructing a power grid geological disaster meteorological risk prediction model according to the first aspect or any of its corresponding embodiments, or the method for medium-term prediction of power grid geological disaster meteorological risk according to the second aspect or any of its corresponding embodiments.

[0039] In a seventh aspect, the present invention provides a computer program product, including computer instructions, which are used to cause a computer to execute the method for constructing a meteorological risk prediction model for power grid geological disasters according to the first aspect or any of its corresponding embodiments, or the method for medium-term prediction of meteorological risk of power grid geological disasters according to the second aspect or any of its corresponding embodiments. Attached Figure Description

[0040] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0041] Figure 1 This is a flowchart illustrating the method for constructing a meteorological risk prediction model for power grid geological disasters according to an embodiment of the present invention. Figure 2 This is a flowchart illustrating the medium-term prediction method for meteorological risks of geological disasters in power grids according to an embodiment of the present invention. Figure 3 This is a schematic diagram of the medium-term (7-day) meteorological risk forecast for a power grid in a certain region according to an embodiment of the present invention. Figure 4 This is a structural block diagram of a power grid geological disaster meteorological risk prediction model construction device according to an embodiment of the present invention; Figure 5This is a structural block diagram of a medium-term forecasting device for meteorological risks of geological disasters in power grids according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0043] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0044] 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 technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0045] According to an embodiment of the present invention, an embodiment of a method for constructing a meteorological risk prediction model for geological disasters in power grids is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0046] This embodiment provides a method for constructing a meteorological risk prediction model for power grid geological disasters. Figure 1 This is a flowchart of a method for constructing a meteorological risk prediction model for power grid geological disasters according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps: Step S101: Obtain historical meteorological data, geological data, tower foundation deformation data, power grid operation data, and corresponding historical disaster data.

[0047] Specifically, meteorological data includes rainfall and temperature data, which can be obtained from sources such as weather forecasts; geological data includes topographic slope and soil / rock type data, which can be obtained through geological department forecasts, geological data, or remote sensing technology. It should be noted that this meteorological and geological data pertains to the region where power grid forecasting is required. Tower foundation deformation data includes data reflecting the structural stability of the tower, such as its horizontal position and vertical settlement; power grid operation data includes data reflecting the equipment's disaster resistance capabilities, such as line load, insulator leakage current, and tower grounding resistance.

[0048] The sampling frequency for meteorological and geological data can be determined based on actual conditions, such as 10 minutes / time or 15 minutes / time, or it can be determined based on the sampling frequency used in meteorological prediction of geological disaster risks using relevant technologies. The sampling frequency for tower foundation deformation data can be 15 minutes / time, and the sampling frequency for power grid operation data is 5 minutes / time to 60 minutes / time. Historical disaster data includes data on geological disasters that occurred in historical times due to meteorological conditions. In addition to actual geological disaster data, this historical disaster data also includes data such as geological disaster meteorological warnings, such as disaster risk warnings jointly issued by geological and meteorological departments based on current geological background, precipitation trends, and other factors.

[0049] It should be noted that the acquired meteorological data, geological data, tower foundation deformation data, and power grid operation data can be time-series data covering a period of time before or after a geological disaster or geological disaster warning. Specifically, the relevant data for the period of time before or after a geological disaster or geological disaster warning can be acquired according to the corresponding sampling frequency, i.e., meteorological time-series data, geological time-series data, tower foundation deformation time-series data, and power grid operation time-series data for the corresponding time period can be acquired.

[0050] Step S102: Extract feature data from the meteorological data, geological data, tower foundation deformation data, and power grid operation data. Specifically, different feature extraction methods can be used for different data, or the same feature extraction method can be used, such as statistical analysis methods, machine learning, or deep learning.

[0051] Step S103: The time-series network model containing tower branches and power grid branches is trained using the feature data to obtain a power grid geological disaster meteorological risk prediction model. The tower branches are used to process the feature data of tower foundation deformation data to obtain processed tower features. The power grid branches are used to process the feature data of power grid operation data to obtain processed power grid features. The time-series network model is used to process the feature data of the meteorological data and geological data, as well as the tower features and the power grid features, to output the risk value for a preset number of days in the future.

[0052] Specifically, since the acquired meteorological data, geological data, tower foundation deformation data, and power grid operation data are all time-series data, this embodiment uses a time-series network model as the basic model. This time-series network model can be a Long Short-Term Memory (LSTM) network model, or it can employ a recurrent neural network or gated recurrent units, etc. Furthermore, this time-series network model includes tower branches and power grid branches, enabling separate processing of the feature data of tower foundation deformation data and power grid operation data. The feature data processed by the two branches are then fused with meteorological and geological characteristics before being input into the time-series network model. Thus, through the pre-processing of the two branches, interference that may arise from directly mixing different types of data in the time-series network model can be avoided, and it can be ensured that the time-series network model does not ignore tower and power grid data.

[0053] This embodiment provides a medium-term prediction method for meteorological risks of geological disasters in power grids, which includes the following steps: Step S201 involves acquiring historical meteorological data, geological data, tower foundation deformation data, power grid operation data, and corresponding historical disaster data. Specifically, after acquiring this multi-source data, preprocessing can be performed before subsequent feature extraction to ensure the quality of the extracted features. The preprocessing process may include cleaning, denoising, and normalization. For missing data, interpolation and machine learning algorithms can be used to fill in the gaps.

[0054] For example, outliers can be removed from tower foundation deformation data using the 3σ principle (Laida criterion). Line load data in power grid operation data can be processed using a 5-point moving average, which uses the average of data within a small window containing the current point and its neighbors to represent the current point's value. This effectively smooths out rapid, random, small fluctuations in the data, making the overall trend clearer. For missing data, linear interpolation can be used to complete short-term missing data (less than 2 hours); while for long-term missing data (2-24 hours), the K-Nearest Neighbor (KNN) algorithm can be used, where each sample is represented by its K nearest neighbors. The value of K can be determined based on the actual situation, for example, K=5.

[0055] Step S202: Extract feature data from the meteorological data, geological data, tower foundation deformation data, and power grid operation data.

[0056] Specifically, step S202 includes: Step S2021 involves extracting feature data from the meteorological and geological data using a preset algorithm. Specifically, meteorological features can be extracted from the meteorological data through statistical processing, threshold processing, and time-series processing. For example, cumulative rainfall, maximum rainfall, average temperature, and highest and lowest temperatures within a preset time period can be statistically analyzed to obtain meteorological statistical features. Thresholds can also be set, such as whether rainfall exceeds a threshold, to obtain meteorological threshold features for exceeding or not exceeding the threshold. Furthermore, the slope of rainfall change within a preset time period can be determined to obtain meteorological time-series features. For geological data, directly extracted data can be used as features, such as slope features, aspect features, and soil and rock type features. Type features can also be obtained through category coding to obtain data feature vectors. The above are merely illustrative examples of feature extraction from meteorological and geological data. In other embodiments, other feature extraction methods, such as machine learning, can be used, and this embodiment does not specifically limit these methods.

[0057] Step S2022: Determine the cumulative deformation of the tower based on the tower foundation deformation data to obtain the characteristic data of the tower foundation deformation data; the cumulative deformation of the tower is determined using the following formula:

[0058] In the formula, This represents the tower foundation deformation data at time t. This represents the tower foundation deformation data at time t-t0. t0 can be determined based on actual conditions; for example, t0=72, meaning the tower foundation deformation data for the three days prior to time t is obtained. Therefore, the cumulative deformation of the tower indicates how much the tower deformation has increased over the past three days.

[0059] Step S2023: Determine the power grid load mutation coefficient based on the line load, and determine the equipment degradation index based on the insulator leakage current and tower grounding resistance to obtain characteristic data of power grid operation data.

[0060] The load mutation coefficient is expressed by the following formula:

[0061] In the formula, This represents the maximum load at time t. This represents the average load at time t; where the maximum load at time t can be the maximum value of the line load at all historical times; the average load at time t can be the average of the line load at time t and the line load within a previous preset time period. For example, the average load can be obtained by using a sliding time window, that is, by using a sliding time window to process the line load at historical times and taking the average value of each sliding time window as the average load.

[0062] The equipment degradation index is determined using the following formula:

[0063] In the formula, This represents the insulator leakage current at time t. This indicates the maximum leakage current of the insulator. This represents the tower grounding resistance at time t. This indicates the maximum value of the tower's grounding resistance. and This represents the preset weights. The maximum insulator leakage current and the maximum tower grounding resistance can be the maximum values ​​of all insulator leakage currents and tower grounding resistances acquired at any given historical moment. Therefore, dividing the value at each moment by the maximum value normalizes the relevant data, placing it within the [0,1] interval for subsequent model processing. Preset weights and It can be determined based on historical data.

[0064] Step S2024: Perform exponentially weighted smoothing on the cumulative deformation of the tower to obtain the processed cumulative deformation. Specifically, this exponentially weighted smoothing means that for the calculated cumulative deformation of the tower at each moment, it can be weighted and summed with the processed cumulative deformation at the previous moment to obtain the processed cumulative deformation at the current moment. Through this exponentially weighted smoothing, small fluctuations in the cumulative deformation sequence at multiple moments can be smoothed out, making the continuous cumulative deformation trend more obvious and smooth.

[0065] Step S2025 involves using principal component analysis (PCA) to filter the feature data, obtaining the filtered features. Specifically, PCA can be used to reduce the dimensionality of the obtained feature data, retaining principal components with a cumulative variance contribution rate ≥90%, thus improving model efficiency. It should be noted that this dimensionality reduction process is primarily for features with a large amount of data. For example, if meteorological data contains many features, PCA is used for dimensionality reduction. However, if only the grid load mutation coefficient and equipment degradation index are extracted from the grid operation data, PCA can be omitted for dimensionality reduction, and the data can be directly input into subsequent grid branches for processing.

[0066] Step S203: The time-series network model containing tower branches and power grid branches is trained using the feature data to obtain a power grid geological disaster meteorological risk prediction model. The tower branches are used to process the feature data of tower foundation deformation data to obtain processed tower features. The power grid branches are used to process the feature data of power grid operation data to obtain processed power grid features. The time-series network model is used to process the feature data of the meteorological data and geological data, as well as the tower features and the power grid features, to output the risk value for a preset number of days in the future.

[0067] The tower branch is used to assign weights to the cumulative deformation of the tower. In this embodiment, an attention network is set in the tower branch. This attention network uses an attention mechanism to assign different attention weights to different deformation amounts, such as strengthening the weight of large cumulative deformation features to highlight the risk of tower collapse. Specifically, the attention network calculates an attention score based on the magnitude of the cumulative deformation, then normalizes the attention score using a normalization function to obtain the corresponding attention weight. Finally, it sums all the cumulative deformation amounts with weights, or generates a weighted cumulative deformation amount as the output of the tower branch.

[0068] The power grid branch is used to correlate the coordinated risks of power grid load mutation coefficient and equipment degradation index. In this embodiment, a neural network is set up in the power grid branch for the interaction between the two features. Specifically, this neural network can be a fully connected neural network, which performs a nonlinear transformation on the two features based on the weight matrix, bias, and activation function, thereby outputting a feature vector after the two features are correlated. This vector integrates the coordinated information of load and equipment status.

[0069] Specifically, the features output from the power grid branches and tower branches are concatenated with meteorological and geological features and then input into the backbone of the time-series network model. In this embodiment, a Long Short-Term Memory (LSTM) network with 128 hidden layer units and a dropout rate of 0.2 is used. This network processes the concatenated features to output daily risk values ​​for a predetermined number of days in the future, such as the next 7 days. Furthermore, during the training of the time-series network model, the acquired historical data can be divided into training and testing sets, and the Adam optimizer (adaptive learning rate optimization algorithm) with a learning rate of 0.001 is used. An early stopping strategy is employed: the model stops if the validation set loss does not decrease for five consecutive rounds to avoid overfitting.

[0070] In addition, the loss function used during training is expressed by the following formula:

[0071] In the formula, Indicates the mean square error loss. This indicates the loss due to tower deformation constraint. Indicates load constraint loss. and This represents the balancing weights. Specifically, this embodiment adds two constraint losses to the mean square error loss function in related technologies. The tower deformation constraint loss is set as an additional loss when the cumulative tower deformation exceeds a safety threshold; that is, when the cumulative tower deformation does not exceed the safety threshold, this loss is 0, and when it does exceed, the greater the excess, the greater the loss. The load constraint loss has a similar meaning to the tower deformation constraint loss; that is, an additional loss when the grid load mutation coefficient exceeds a mutation threshold; when the grid load mutation coefficient does not exceed the mutation threshold, this loss is 0, and when it does exceed, the greater the excess, the greater the loss. The two balancing weights can be determined based on actual conditions or historical data. In this embodiment, λ 1 = 0.3 λ 2 = 0.2.

[0072] Furthermore, the constructed power grid geological disaster meteorological risk prediction model can be evaluated using indicators such as cross-validation, mean squared error, and accuracy to ensure its prediction accuracy and stability. Based on the evaluation results, model parameters can be adjusted or new features can be introduced to further improve the model's predictive performance.

[0073] As one or more specific application embodiments of the present invention, the method for constructing a meteorological risk prediction model for power grid geological disasters is implemented using the following process: (a) Data acquisition and preprocessing.

[0074] (1) Data source: This invention uses multi-source data, including: meteorological forecasts, geological data, historical disaster data, tower foundation deformation data and power grid operation data required for traditional geological disaster meteorological risk prediction, as shown in Table 1.

[0075] Table 1. Data Sources for Meteorological Risk Prediction of Geological Disasters in Power Grids

[0076] (2) Specialized preprocessing: The collected multi-source data is cleaned, denoised, and normalized to ensure data quality. For missing data, interpolation or machine learning algorithms are used to fill in the missing data.

[0077] 1) Tower foundation deformation data: Outliers were removed using the 3σ principle (Laida criterion).

[0078] 2) Power grid operation data: Load fluctuations are handled using a 5-point moving average.

[0079] 3) Missing data: For short-term missing data less than 2 hours, linear interpolation is used; for long-term missing data of 2-24 hours, the K-nearest neighbor algorithm (KNN algorithm) is used to fill in the missing data (K=5).

[0080] (ii) Feature engineering.

[0081] (1) Feature extraction: Extract specific features from the preprocessed data.

[0082] Calculate the cumulative deformation (formula: Then, noise reduction is achieved through exponential weighted smoothing.

[0083] Construct the equipment degradation index (formula: ).

[0084] Calculate the power grid load mutation coefficient (formula: .

[0085] (2) Feature selection: Principal component analysis is used to reduce dimensionality and retain principal components with a cumulative variance contribution rate of ≥90% to improve model efficiency.

[0086] (III) Model building and training.

[0087] (1) Model structure: A power grid geological disaster meteorological risk prediction model is constructed, with the Long Short-Term Memory Network (LSTM) model as the backbone and two new dedicated branches: the tower deformation attention branch and the power grid operation correlation branch.

[0088] 1) Tower Deformation Attention Branch: Strengthen the weight of large cumulative deformation characteristics to highlight the risk of tower collapse.

[0089] 2) Power grid operation related branches: Associate load changes and equipment degradation risks to improve the risk capture capability of heavy-load lines; after splicing the branch features, input them into LSTM (128 hidden layer units, dropout rate=0.2) to output the daily risk value (0-10) for the next 7 days.

[0090] (2) Constraint loss function: Optimize the traditional mean square error (MSE) loss by adding power facility state constraints, the formula is: ,in, This is the loss due to mean squared error (MSE). To constrain tower deformation, losses are increased when the deformation exceeds the safety threshold; To constrain the load, the loss is increased when the threshold for mutation is exceeded; λ 1 = 0.3 λ 2 = 0.2) is the balancing weight. For example, the safety threshold is 5mm, and the mutation threshold is 0.2.

[0091] (3) Training simplification: Data from the past 3 years (including 32 geological disaster records) were selected and divided into training and test datasets in a 7:3 ratio; the Adam optimizer (adaptive learning rate optimization algorithm) was used with a learning rate of 0.001; an early stopping strategy was adopted, and the test dataset was stopped if the loss on the validation set did not decrease for 5 consecutive rounds to avoid overfitting.

[0092] (iv) Model evaluation and optimization.

[0093] (1) Model evaluation: The model is evaluated using indicators such as cross-validation, mean squared error (MSE), and accuracy to ensure the model's prediction accuracy and stability.

[0094] (2) Model optimization: Based on the evaluation results, adjust the model parameters or introduce new features to further improve the model's predictive performance.

[0095] This embodiment also provides a method for medium-term prediction of meteorological risks of geological disasters in power grids, such as... Figure 2 As shown, the method includes: Step S301: Acquire meteorological data, geological data, tower foundation deformation data, and power grid operation data. Specifically, the acquired meteorological data can be the meteorological department's forecast of the next seven days, such as the forecast data for the next seven days. Geological data, tower foundation deformation data, and power grid operation data, due to their inertia and periodicity, change relatively little in a short period of time, and therefore can be acquired in real time.

[0096] Step S302: Extract feature data from the meteorological data, geological data, tower foundation deformation data, and power grid operation data. Specifically, when extracting feature data, the feature extraction method of step S202 above can be referred to, and the corresponding formulas can be used to calculate the cumulative deformation of the tower, the power grid load mutation coefficient, and the equipment degradation index. In the calculation, the maximum values ​​in the formula, such as the maximum load, the maximum leakage current of the insulator, and the maximum grounding resistance of the tower, can be directly adopted from the values ​​determined during the training process.

[0097] Step S303: Input the feature data into the power grid geological disaster meteorological risk prediction model constructed by the power grid geological disaster meteorological risk prediction model construction method described in the above embodiment to obtain the risk value for the next preset number of days. Specifically, after extracting the feature data, it is input into the power grid geological disaster meteorological risk prediction model constructed during the training process to obtain the daily risk value for the next preset number of days (e.g., the next seven days).

[0098] In an optional implementation, the method further includes: determining a risk level based on the risk value of the preset number of future days; and generating a handling result for the power grid and towers based on the risk level. Specifically, for the determined risk value, a predetermined range can be used to determine its corresponding risk level. The classification and expression of meteorological risk levels for power grid geological disasters in this embodiment are shown in Table 2 below: Table 2 Classification and Expression of Meteorological Risk Levels for Geological Disasters in Power Grids

[0099] It should be noted that the range for each risk level can be determined based on the actual situation. After determining the risk level, corresponding measures can be taken. For example, if the risk level is Level III, on-site inspections and foundation grouting reinforcement measures were carried out for the high-risk towers.

[0100] As one or more specific application embodiments of the present invention, the medium-term prediction method for meteorological risks of geological disasters in power grids is implemented using the following process: (1) Data acquisition: Maximum cumulative deformation of the tower (5mm safety threshold), power grid load mutation coefficient (0.2 threshold) The rainfall in the next 7 days is 85mm.

[0101] (2) Prediction results: The model outputs a risk value of 4.6, which is classified as a Level III (yellow) warning according to the three-dimensional matrix. The prediction results are as follows. Figure 3 As shown.

[0102] (3) Response effect: After the warning was issued, the regional power grid conducted on-site inspections and foundation grouting reinforcement of high-risk towers. No tilting or line interruption occurred after the rainfall, thus avoiding line outage losses.

[0103] This invention, for the first time, integrates tower foundation deformation data with power grid operation data, filling the gap in the correlation between "facility status and disaster risk" in medium-term forecasting. Simultaneously, a specific processing flow and model optimization were designed for tower foundation deformation and power grid operation data, namely, constructing power grid branches and tower branches, reducing false alarm and false alarm rates. Furthermore, a method for predicting the meteorological risk of power grid geological disasters for the next 7 days is proposed, providing a scientific basis for power grid disaster prevention and mitigation.

[0104] This embodiment also provides a device for constructing a meteorological risk prediction model for power grid geological disasters. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0105] This embodiment provides a device for constructing a meteorological risk prediction model for geological disasters in power grids, such as... Figure 4 As shown, it includes: The historical data acquisition module 41 is used to acquire historical meteorological data, geological data, tower foundation deformation data, power grid operation data, and corresponding historical disaster data. Historical feature extraction module 42 is used to extract feature data from the meteorological data, geological data, tower foundation deformation data and power grid operation data; The model training and construction module 43 is used to train a time-series network model containing tower branches and power grid branches using the feature data to obtain a power grid geological disaster meteorological risk prediction model. The tower branches are used to process the feature data of tower foundation deformation data to obtain processed tower features. The power grid branches are used to process the feature data of power grid operation data to obtain processed power grid features. The time-series network model is used to process the feature data of the meteorological data and geological data, as well as the tower features and the power grid features, to output a risk value for a preset number of days in the future.

[0106] In one optional implementation, the power grid operation data includes line load, insulator leakage current, and tower grounding resistance; the historical feature extraction module is specifically used to extract feature data from the meteorological and geological data using a preset algorithm; the cumulative deformation of the tower is determined based on the tower foundation deformation data to obtain feature data of the tower foundation deformation data; the power grid load mutation coefficient is determined based on the line load, and the equipment degradation index is determined based on the insulator leakage current and tower grounding resistance to obtain feature data of the power grid operation data.

[0107] In one optional implementation, the cumulative deformation of the tower is determined using the following formula:

[0108] In the formula, This represents the tower foundation deformation data at time t. This represents the tower foundation deformation data at time t-t0; The load mutation coefficient is expressed by the following formula:

[0109] In the formula, This represents the maximum load at time t. This represents the average load at time t; The equipment degradation index is determined using the following formula:

[0110] In the formula, This represents the insulator leakage current at time t. This indicates the maximum leakage current of the insulator. This represents the tower grounding resistance at time t. This indicates the maximum value of the tower's grounding resistance. and This indicates the preset weight.

[0111] In one optional implementation, the historical feature extraction module is specifically used to perform exponential weighted smoothing on the cumulative deformation of the tower to obtain the processed cumulative deformation; and to use principal component analysis to filter the feature data to obtain the filtered features.

[0112] In one alternative implementation, the loss function used during training is expressed by the following formula:

[0113] In the formula, Indicates the mean square error loss. This indicates the loss due to tower deformation constraint. Indicates load constraint loss. and This indicates the balancing weight.

[0114] The power grid geological disaster meteorological risk prediction model construction method provided in this embodiment of the invention can execute the power grid geological disaster meteorological risk prediction model construction method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method. Further functional descriptions of the above modules and units are the same as in the corresponding embodiments above, and will not be repeated here.

[0115] This embodiment also provides a medium-term prediction device for meteorological risks of geological disasters in power grids. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0116] This embodiment provides a medium-term prediction device for meteorological risks of geological disasters in power grids, such as... Figure 5 As shown, it includes: Data acquisition module 51 is used to acquire meteorological data, geological data, tower foundation deformation data, and power grid operation data; Feature extraction module 52 is used to extract feature data from the meteorological data, geological data, tower foundation deformation data and power grid operation data; Prediction module 53 is used to input the feature data into the power grid geological disaster meteorological risk prediction model constructed by the power grid geological disaster meteorological risk prediction model construction method described in the above embodiment, and obtain the risk value for a preset number of days in the future.

[0117] In one optional embodiment, the device further includes: a risk level determination module, configured to determine a risk level based on the risk value of the preset number of future days; and a handling module, configured to generate handling results for the power grid and towers based on the risk level.

[0118] The medium-term forecasting device for meteorological risks of geological disasters in power grids provided in this embodiment of the invention can execute the medium-term forecasting method for meteorological risks of geological disasters in power grids provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0119] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0120] The following is a detailed reference. Figure 6This diagram illustrates a structural schematic suitable for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, a graphics processing unit, etc.) 11, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 12 or a program loaded from memory 18 into random access memory (RAM) 13. The RAM 13 also stores various programs and data required for the operation of the electronic device. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0121] Typically, the following devices can be connected to I / O interface 15: input devices 16 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 17 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 18 including, for example, magnetic tapes, hard disks, etc.; and communication devices 19. Communication device 19 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 6 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0122] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 19, or installed from a memory 18, or installed from a ROM 12. When the computer program is executed by the processor 11, it performs the functions defined in the medium-term prediction method for meteorological risks of power grid geological disasters according to embodiments of the present invention.

[0123] Figure 6 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0124] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the medium-term prediction method for meteorological risks of power grid geological disasters shown in the above embodiments is implemented.

[0125] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0126] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and all such modifications and variations fall within the scope defined by the appended claims.

Claims

1. A method for constructing a meteorological risk prediction model for geological disasters in power grids, characterized in that, The method includes: Acquire historical meteorological data, geological data, tower foundation deformation data, power grid operation data, and corresponding historical disaster data; Extract feature data from the meteorological data, geological data, tower foundation deformation data, and power grid operation data; The time-series network model, which includes tower branches and power grid branches, is trained using the aforementioned feature data to obtain a power grid geological disaster meteorological risk prediction model. The tower branches are used to process the feature data of tower foundation deformation data to obtain processed tower features, and the power grid branches are used to process the feature data of power grid operation data to obtain processed power grid features. The time-series network model is used to process the feature data of the meteorological and geological data, as well as the tower features and the power grid features, to output a risk value for a preset number of days in the future.

2. The method according to claim 1, characterized in that, The power grid operation data includes line load, insulator leakage current, and tower grounding resistance; The feature data extracted from the meteorological data, geological data, tower foundation deformation data, and power grid operation data include: Feature data of the meteorological and geological data are extracted using a preset algorithm; Based on the tower foundation deformation data, the cumulative deformation of the tower is determined, and the characteristic data of the tower foundation deformation data is obtained. Based on the line load, the power grid load mutation coefficient is determined, and based on the insulator leakage current and tower grounding resistance, the equipment degradation index is determined, thus obtaining characteristic data of the power grid operation data.

3. The method according to claim 2, characterized in that, The cumulative deformation of the tower is determined using the following formula: In the formula, This represents the tower foundation deformation data at time t. This represents the tower foundation deformation data at time t-t0; The load mutation coefficient is expressed by the following formula: In the formula, This represents the maximum load at time t. This represents the average load at time t; The equipment degradation index is determined using the following formula: In the formula, This represents the insulator leakage current at time t. This indicates the maximum leakage current of the insulator. This represents the tower grounding resistance at time t. This indicates the maximum value of the tower's grounding resistance. and This indicates the preset weight.

4. The method according to claim 2, characterized in that, The extraction of feature data from the meteorological data, geological data, tower foundation deformation data, and power grid operation data also includes: The cumulative deformation of the tower is subjected to exponential weighted smoothing to obtain the processed cumulative deformation. Principal component analysis was used to filter the feature data, resulting in the filtered features.

5. The method according to claim 1, characterized in that, The loss function used during training is expressed by the following formula: In the formula, Indicates the mean square error loss. This indicates the loss due to tower deformation constraint. Indicates load constraint loss. and This indicates the balancing weight.

6. The method according to claim 2, characterized in that, The tower branch is used to assign weights to the cumulative deformation of the tower, and the power grid branch is used to correlate the synergistic risks of the power grid load mutation coefficient and the equipment degradation index.

7. A method for medium-term prediction of meteorological risk of geological disasters in power grids, characterized in that, The method includes: Acquire meteorological data, geological data, tower foundation deformation data, and power grid operation data; Extract feature data from the meteorological data, geological data, tower foundation deformation data, and power grid operation data; The feature data is input into the power grid geological disaster meteorological risk prediction model constructed by the power grid geological disaster meteorological risk prediction model construction method according to any one of claims 1-6 to obtain the risk value for the next preset number of days.

8. The method according to claim 7, characterized in that, The method further includes: The risk level is determined based on the risk value of the predetermined number of future days; The appropriate action plan for the power grid and power poles is generated based on the risk level.

9. A device for constructing a meteorological risk prediction model for power grid geological disasters, characterized in that, The method includes: The historical data acquisition module is used to acquire historical meteorological data, geological data, tower foundation deformation data, power grid operation data, and corresponding historical disaster data. The historical feature extraction module is used to extract feature data from the meteorological data, geological data, tower foundation deformation data, and power grid operation data. The model training and construction module is used to train a time-series network model containing tower branches and power grid branches using the feature data to obtain a power grid geological disaster meteorological risk prediction model. The tower branches are used to process the feature data of tower foundation deformation data to obtain processed tower features, and the power grid branches are used to process the feature data of power grid operation data to obtain processed power grid features. The time-series network model is used to process the feature data of the meteorological data and geological data, as well as the tower features and the power grid features, to output a risk value for a preset number of days in the future.

10. The apparatus according to claim 9, characterized in that, The power grid operation data includes line load, insulator leakage current, and tower grounding resistance. The historical feature extraction module is specifically used to extract feature data from the meteorological and geological data using a preset algorithm. Based on the tower foundation deformation data, the cumulative deformation of the tower is determined to obtain the feature data of the tower foundation deformation data. Based on the line load, the power grid load mutation coefficient is determined, and based on the insulator leakage current and tower grounding resistance, the equipment degradation index is determined to obtain the feature data of the power grid operation data.

11. The apparatus according to claim 10, characterized in that, The cumulative deformation of the tower is determined using the following formula: In the formula, This represents the tower foundation deformation data at time t. This represents the tower foundation deformation data at time t-t0; The load mutation coefficient is expressed by the following formula: In the formula, This represents the maximum load at time t. This represents the average load at time t; The equipment degradation index is determined using the following formula: In the formula, This represents the insulator leakage current at time t. This indicates the maximum leakage current of the insulator. This represents the tower grounding resistance at time t. This indicates the maximum value of the tower's grounding resistance. and This indicates the preset weight.

12. The apparatus according to claim 10, characterized in that, The historical feature extraction module is specifically used to perform exponential weighted smoothing on the cumulative deformation of the tower to obtain the processed cumulative deformation; principal component analysis is used to filter the feature data to obtain the filtered features.

13. The apparatus according to claim 9, characterized in that, The loss function used during training is expressed by the following formula: In the formula, Indicates the mean square error loss. This indicates the loss due to tower deformation constraint. Indicates load constraint loss. and This indicates the balancing weight.

14. The apparatus according to claim 9, characterized in that, The tower branch is used to assign weights to the cumulative deformation of the tower, and the power grid branch is used to correlate the synergistic risks of the power grid load mutation coefficient and the equipment degradation index.

15. A medium-term prediction device for meteorological risks of geological disasters in power grids, characterized in that, The device includes: The data acquisition module is used to acquire meteorological data, geological data, tower foundation deformation data, and power grid operation data; The feature extraction module is used to extract feature data from the meteorological data, geological data, tower foundation deformation data, and power grid operation data; The prediction module is used to input the feature data into the power grid geological disaster meteorological risk prediction model constructed by the power grid geological disaster meteorological risk prediction model construction method according to any one of claims 1-6, and obtain the risk value for a preset number of days in the future.

16. The apparatus according to claim 15, characterized in that, The device further includes: The risk level determination module is used to determine the risk level based on the risk value of the preset number of future days. The handling module is used to generate handling results for the power grid and towers based on the risk level.

17. An electronic device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the power grid geological disaster meteorological risk prediction model construction method as described in any one of claims 1 to 6, or the power grid geological disaster meteorological risk mid-term prediction method as described in claim 7 or 8.

18. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, which are used to cause the computer to execute the power grid geological disaster meteorological risk prediction model construction method according to any one of claims 1 to 6 or to execute the power grid geological disaster meteorological risk mid-term prediction method according to claim 7 or 8.

Images