Method, device and equipment for predicting icing thickness of overhead transmission line and medium

Abstract

The invention relates to the technical field of power grid safety, and discloses an overhead transmission line icing thickness prediction method and device, equipment and a medium. The method comprises the following steps: acquiring historical data and future meteorological data of a to-be-predicted overhead line; performing preprocessing and feature engineering on the historical data to obtain optimized historical data; dividing the optimized historical data into a training set and a test set, training the plurality of weak learner candidate models through the training set, calculating the score of each weak learner candidate model on the verification set, and determining a strong learner model; and dividing the optimized historical data and the future meteorological data into a total verification set and a total test set, training a meta-learner model through each strong learner model and the total verification set, and inputting a prediction result of each strong learner model on the total test set into the meta-learner model to obtain a final icing thickness prediction value of the total test set. According to the method, the future icing thickness is predicted through multi-model fusion, and the accuracy of icing thickness prediction of the overhead transmission line is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid security, and particularly to a method, device, equipment and medium for predicting the icing thickness of overhead transmission lines. Background Art

[0002] The prediction of icing on overhead lines is an important technical means for disaster prevention and mitigation in the power system. The aim is to predict the possible icing phenomenon of transmission lines under meteorological conditions such as low temperature and high humidity through scientific methods, so as to take protective measures in advance and ensure the safe and stable operation of the power grid. Icing can lead to an increase in the mechanical load of the conductor, an increase in the sag, and even serious accidents such as wire breakage and tower collapse. Therefore, icing prediction is of great significance to the reliability and economy of the power system.

[0003] At present, most of the existing methods for predicting icing on overhead lines have poor model prediction and generalization capabilities, which restricts the accuracy of predicting the icing thickness in the actual production environment and cannot well solve the prediction problems under conditions such as nonlinear and large random fluctuations of historical training data. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to overcome the deficiencies in the prior art and provide a method, device, equipment and medium for predicting the icing thickness of overhead transmission lines.

[0005] The present invention provides the following technical solutions:

[0006] In the first aspect, the present invention provides a method for predicting the icing thickness of an overhead transmission line, the method comprising:

[0007] Obtain the historical data and future meteorological data of the overhead line to be predicted, where the historical data includes historical meteorological data, line data, equipment data and historical icing thickness data;

[0008] Preprocess the historical data to obtain preprocessed historical data, and perform feature engineering on the preprocessed historical data to obtain optimized historical data;

[0009] Divide the optimized historical data into a training set and a test set, where the training set contains multiple validation sets, train multiple weak learner candidate models through the training set, calculate the scores of each weak learner candidate model on the validation set, and determine multiple strong learner models according to the scores of each weak learner candidate model;

[0010] The optimized historical data and the future meteorological data are combined into a data set, and the data set is divided into a total validation set and a total test set. A meta-learner model is trained by each of the strong learner models and the total validation set, and the prediction results of each of the strong learner models on the total test set are input into the meta-learner model to obtain the final ice thickness prediction value of the total test set.

[0011] In an optional implementation manner, the preprocessing the historical data to obtain the preprocessed historical data includes:

[0012] Sampling and aligning the historical data at preset time intervals to obtain aligned historical data;

[0013] Determine missing time period data in the aligned historical data, and perform mean filling on the missing time period data to obtain filled historical data;

[0014] Determine abnormal data in the filled historical data, and remove the abnormal data to obtain the removed historical data;

[0015] Determine the missing historical meteorological data in the eliminated historical data, and perform Nan value filling on the missing historical meteorological data to obtain the preprocessed historical data.

[0016] In an optional implementation manner, performing feature engineering on the preprocessed historical data to obtain optimized historical data includes:

[0017] Normalizing the historical meteorological data and the historical ice thickness data to obtain normalized historical data;

[0018] Performing one-hot encoding on the line data and the device data to obtain one-hot encoded historical data;

[0019] Principal component analysis is performed on the one-hot encoded historical data, noise field columns in the one-hot encoded historical data are deleted, and data dimension reduction is performed on the one-hot encoded historical data to obtain the optimized historical data.

[0020] In an optional implementation, the optimization historical data is divided into a training set and a test set, and the training set contains multiple validation sets, including:

[0021] Dividing the optimized historical data into the training set and the test set according to a preset ratio;

[0022] The training set is divided into multiple groups of first training sets by performing K-fold cross-division using a Bagging algorithm;

[0023] Divide each of the first training sets into a validation set and multiple second training sets.

[0024] In an alternative embodiment, calculating the scores of the candidate weak learner models on the validation set and determining multiple strong learner models based on the scores of the candidate weak learner models includes:

[0025] Predict the validation set through each of the candidate weak learner models to obtain the predicted icing thickness values of each of the candidate weak learner models on the validation set;

[0026] Obtain the actual icing thickness values of the validation set, and obtain the number of icing lines and the number of time series segments of the data in the validation set;

[0027] Based on the predicted icing thickness values, predicted icing thickness values, number of icing lines, and number of time series segments of the validation set, and using a preset model score calculation formula to calculate the scores of each of the candidate weak learner models;

[0028] Arrange the scores of each of the candidate weak learner models in ascending order, and determine the first preset number of candidate weak learner models as the multiple strong learner models.

[0029] In an alternative embodiment, the preset model score calculation formula is:

[0030]

[0031] In the formula, MAPE is the score of the candidate weak learner model, R is the number of icing lines of the data in the validation set, T is the number of time series segments of the data in the validation set, d rt is the actual icing thickness value of the r-th line and t-th time period in the validation set, p rt is the predicted icing thickness value of the r-th line and t-th time period in the validation set.

[0032] In an alternative embodiment, training the meta-learner model through each of the strong learner models and the total validation set includes:

[0033] Predict the total validation set through each of the strong learner models to obtain the predicted icing thickness values of each of the strong learner models on the total validation set;

[0034] Obtain the actual icing thickness values of the total validation set, use the predicted icing thickness values of the total validation set as the feature columns, and use the actual icing thickness values of the total validation set as the target columns to form a total training set, and train the meta-learner model through the total training set.

[0035] In a second aspect, the present invention provides an icing thickness prediction device for an overhead transmission line, the device comprising:

[0036] An acquisition module, configured to acquire historical data and future meteorological data of an overhead line to be predicted, where the historical data includes historical meteorological data, line data, equipment data, and historical icing thickness data;

[0037] A processing module, configured to preprocess the historical data to obtain preprocessed historical data, and perform feature engineering on the preprocessed historical data to obtain optimized historical data;

[0038] A training module, configured to divide the optimized historical data into a training set and a test set, where the training set contains multiple validation sets, train multiple weak learner candidate models through the training set, calculate the scores of each weak learner candidate model on the validation set, and determine multiple strong learner models according to the scores of each weak learner candidate model;

[0039] A prediction module, configured to merge the optimized historical data and the future meteorological data into a data set, divide the data set into a total validation set and a total test set, train a meta-learner model through each strong learner model and the total validation set, and input the prediction results of each strong learner model on the total test set into the meta-learner model to obtain the final icing thickness prediction value of the total test set.

[0040] In a third aspect, an embodiment of the present disclosure provides a computer device, the computer device comprising a memory and a processor, where the memory stores a computer program, and when the processor executes the computer program, the steps of the overhead transmission line icing thickness prediction method described in the first aspect are implemented.

[0041] In a fourth aspect, an embodiment of the present disclosure provides a computer-readable storage medium, where the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the overhead transmission line icing thickness prediction method described in the first aspect are implemented.

[0042] Advantages of the present application:

[0043] The ice thickness prediction method for overhead transmission lines provided by the embodiments of the present application obtains the historical data and future meteorological data of the overhead line to be predicted. The historical data includes historical meteorological data, line data, equipment data, and historical ice thickness data. Preprocess the historical data to obtain the preprocessed historical data, and perform feature engineering on the preprocessed historical data to obtain optimized historical data. Divide the optimized historical data into a training set and a test set. The training set contains multiple validation sets. Train multiple weak learner candidate models through the training set, calculate the scores of each weak learner candidate model on the validation set, and determine multiple strong learner models according to the scores of each weak learner candidate model. Merge the optimized historical data and the future meteorological data into a data set, and divide the data set into a total validation set and a total test set. Train a meta-learner model through each strong learner model and the total validation set, and input the prediction results of each strong learner model on the total test set into the meta-learner model to obtain the final ice thickness prediction value of the total test set. The present application predicts the future short-term ice thickness time series through multi-model fusion. By fusing and integrating multiple weak learners, the learning ability and generalization ability for ice thickness data are enhanced. At the same time, short-term meteorological data for the future prediction time period is introduced. By comprehensively considering the covariate factors that affect the ice thickness evolution in history and the future, the ability of the system to capture the temporal characteristics of ice thickness evolution is improved, and the accuracy of ice thickness prediction for overhead transmission lines is improved.

[0044] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following specifically enumerates preferred embodiments and, in conjunction with the accompanying drawings, makes the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts. In each drawing, similar components are numbered similarly.

[0046] Figure 1 Shows a flowchart of an ice thickness prediction method for an overhead transmission line provided by an embodiment of the present application;

[0047] Figure 2 Shows a flowchart of another ice thickness prediction method for an overhead transmission line provided by an embodiment of the present application;

[0048] Figure 3The flowchart of yet another method for predicting the ice coating thickness of an overhead transmission line provided by an embodiment of the present application is shown;

[0049] Figure 4 The curve comparison diagram of the prediction absolute errors of different prediction strategies provided by an embodiment of the present application is shown;

[0050] Figure 5 The structural schematic diagram of an ice coating thickness prediction device for an overhead transmission line provided by an embodiment of the present application is shown;

[0051] Figure 6 The structural schematic diagram of a computer device provided by an embodiment of the present application is shown. Detailed implementation manners

[0052] The embodiments of the present invention will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary only for explaining the present invention and should not be construed as limiting the present invention.

[0053] It should be noted that the terms "first" and "second" are used for descriptive purposes only and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present invention, "a plurality" means two or more unless otherwise specifically defined.

[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs. The terms used in the description of the present specification herein are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the associated listed items.

[0055] Embodiment 1

[0056] As Figure 1 shown, it is the flowchart of a method for predicting the ice coating thickness of an overhead transmission line in an embodiment of the present application. The method for predicting the ice coating thickness of an overhead transmission line provided by the embodiment of the present application includes the following steps:

[0057] Step S110, obtaining historical data and future meteorological data of the overhead line to be predicted, where the historical data includes historical meteorological data, line data, equipment data, and historical ice coating thickness data.

[0058] In this embodiment, historical data and future meteorological data of the overhead line to be predicted in the area to be predicted are first obtained. Among them, the historical data includes historical meteorological data (fields include "time", "rainfall", "air pressure", "humidity", "temperature", "wind direction", "wind speed", etc.), line data (fields include "line ID", "line code", "line name", "voltage level", "city name", etc.), equipment data (fields include "device code", "tower number", "tower code", "operation and maintenance unit", "phase", "manufacturer", "conductor radius", "conductor density", "design thickness", "reference value", "tensile value", etc.) and historical ice coating thickness data sampled at an hourly level.

[0059] By obtaining the historical data and future meteorological data of the overhead line to be predicted, the above method ensures that the prediction model can comprehensively consider various influencing factors, improving the accuracy and comprehensiveness of the prediction. These data provide a solid foundation for subsequent preprocessing, feature engineering, and model training, and are the key guarantee for the performance of the prediction model.

[0060] Step S120: Preprocess the historical data to obtain the preprocessed historical data, and perform feature engineering on the preprocessed historical data to obtain optimized historical data.

[0061] Understandably, after obtaining the historical data of the overhead line to be predicted, in order to ensure the data quality and make it more suitable for subsequent model training, the historical data needs to be preprocessed. The specific process is as follows:

[0062] (1) Since the data may come from different sensors or data sources, the sampling frequencies may be inconsistent. To unify the data format, it is necessary to sample-align the historical data at a preset time interval (such as every hour) to obtain the aligned historical data;

[0063] (2) After the sampling alignment, for the still missing time period data, the method of mean filling is used to obtain the filled historical data. This means that for a missing time period, the average value of the adjacent hourly data before and after it can be calculated and used to fill the missing data. If there is no data before and after the missing hour (for example, at the beginning or end of the dataset), it may be necessary to consider using the mean value of the entire dataset or other appropriate filling strategies;

[0064] (3) Abnormal data may be generated due to reasons such as sensor failures and data entry errors. For the ice thickness data, abnormal data may have a negative impact on model training. Therefore, it is necessary to remove these abnormal data to obtain the historical data after removal. The methods for removing abnormal data may include: using statistical methods (such as the 3σ principle) to identify and remove outliers, using visualization tools such as box plots to identify outliers, setting reasonable thresholds based on domain knowledge, and removing outliers beyond the thresholds.

[0065] (4) For the missing historical meteorological data in the historical data after removal, the method of filling with Nan values can be adopted to obtain the preprocessed historical data. This means that for a missing historical meteorological data point, it is marked as NaN (Not a Number) and processed during subsequent feature engineering or model training.

[0066] Furthermore, in order to convert the original preprocessed historical data into features that can be understood by the subsequent model, it is necessary to perform feature engineering on the preprocessed historical data. The specific process is as follows:

[0067] (1) For the historical meteorological data and historical ice thickness data, since their value ranges may vary greatly (for example, the value ranges of temperature and ice thickness may be completely different), it is necessary to normalize them to convert data with different dimensions into the same scale to obtain the normalized historical data. The methods of normalization may include min-max normalization, Z-score standardization, etc.;

[0068] (2) For the line data and equipment data, since they are categorical variables and their values are discrete, it is necessary to perform one-hot encoding on them to convert the categorical variables into numerical variables to obtain the one-hot encoded historical data. This means that for each categorical variable, a new binary feature column is created. If a sample belongs to a certain category of the categorical variable, the corresponding feature column is 1, otherwise it is 0;

[0069] (3) Principal Component Analysis (PCA) is a commonly used method for data dimensionality reduction. It reduces the dimensionality of data by calculating the principal components of the data (i.e., the main directions of data variation). In the prediction of overhead line icing, due to having a large number of feature columns (including meteorological data, line data, and equipment data, etc.), some of these feature columns may be redundant or contain noise. Therefore, perform principal component analysis on the one-hot encoded historical data to delete these redundant or noisy feature columns and perform data dimensionality reduction processing to obtain the final optimized historical data.

[0070] The above method effectively improves the quality and integrity of data and reduces prediction errors caused by data loss or anomalies through methods such as sampling alignment, mean filling, outlier removal, and Nan value filling. Normalization converts data with different dimensions into the same scale, facilitating model processing; one-hot encoding converts categorical variables into numerical variables, enhancing the model's expressive ability; PCA dimensionality reduction removes redundant and noisy features, simplifies the model structure, and improves computational efficiency.

[0071] Step S130: Divide the optimized historical data into a training set and a test set. The training set contains multiple validation sets. Train multiple weak learner candidate models with the training set, calculate the scores of each weak learner candidate model on the validation set, and determine multiple strong learner models based on the scores of each weak learner candidate model.

[0072] Understandably, as Figure 2 shown, first, divide the optimized historical data into a training set and a test set according to a preset ratio (e.g., 9:1). Use the Bagging algorithm to evenly divide the training set into M groups (e.g., M = 5) of the first training sets by K-fold cross-validation (e.g., K = 5), and then divide each group of the first training sets into a validation set and multiple second training sets. Then, use the M groups of the first training sets to train all weak learner candidate models. In this embodiment, the weak learner candidate models may include XGBoost, LightGBM, SVR, linear regression model, KNNR, Chronos, etc., and the present application does not limit this.

[0073] Next, use all the trained weak learner candidate models to predict the validation set to obtain the predicted ice accretion thickness values of each weak learner candidate model on the validation set. Then, according to the actual ice accretion thickness values, the number of ice-covered lines, and the number of time series segments of the obtained validation set, calculate the scores of each weak learner candidate model through a preset model score calculation formula. In the present application, the scores of each weak learner candidate model are represented by the Mean Absolute Percent Error (MAPE). The specific calculation formula is:

[0074]

[0075] In the formula, MAPE is the score of the weak learner candidate model, R is the number of ice-covered lines in the data of the validation set, T is the number of time series segments in the data of the validation set, d rt is the actual ice accretion thickness value of the r-th line at the t-th time period in the validation set, and p rt is the predicted ice accretion thickness value of the r-th line at the t-th time period in the validation set.

[0076] Further, the scores of each weak learner candidate model are arranged in ascending order. The smaller the score, the better the performance of the model. Therefore, the top preset number (e.g., 3) of weak learner candidate models are selected and determined as multiple strong learner models.

[0077] The above method trains multiple weak learner candidate models and calculates scores using the validation set, ensuring the diversity and complementarity of the models, and providing rich choices for subsequent model fusion. Evaluating the weak learners through evaluation metrics such as MAPE can accurately reflect the prediction performance of the models and provide a reliable basis for determining the strong learner models.

[0078] Step S140: Combine the optimized historical data and the future meteorological data into a data set, and divide the data set into a total validation set and a total test set. Train a meta-learner model using each of the strong learner models and the total validation set, and input the prediction results of each strong learner model on the total test set into the meta-learner model to obtain the predicted icing thickness value of the total test set.

[0079] Understandably, as Figure 3 shown, finally, the optimized historical data and the future meteorological data are fused into a data set, and the data set is divided into a total validation set and a total test set. The multiple strong learner models obtained through step S130 are used to predict the total validation set to obtain the predicted icing thickness value of the total validation set. The predicted icing thickness value of the total validation set is used as the feature column, and the actual icing thickness value of the obtained total validation set is used as the target column. The target column and the feature column are recombined into a total training set, and a meta-model (usually a simple regression model, such as a linear regression model) is trained using this total training set.

[0080] The task of the meta-model is to learn how to fuse the prediction results of the strong learner models and generate the final prediction result. The specific steps are as follows: Use the multiple strong learner models to predict the total test set to obtain the predicted icing thickness values of each strong learner model on the total test set, and input the predicted icing thickness values of each strong learner model on the total test set into the meta-learner model to obtain the final predicted icing thickness value of the total test set.

[0081] The above method combines the Stacking technique. By combining the prediction results of multiple strong learner models, a meta-model is trained to generate the final prediction. The core idea is to utilize the diversity of different models to capture different patterns in the data, thereby improving the overall performance and the accuracy and robustness of the prediction. By comprehensively considering historical and future influencing factors, as well as the prediction results of multiple models, the system can more comprehensively and accurately predict the icing thickness of overhead transmission lines.

[0082] In an alternative embodiment, as Figure 4 shown, it is a comparison chart of the prediction absolute error curves under three different prediction strategies adopted on a certain line (including: ① strong learner model, ② strong learner model + Chronos, ③ strong learner model + Chronos + future meteorological data). It can be seen that the strong learner model + Chronos (weak learner) + future meteorological data is the prediction strategy adopted in this application, and the error value is the smallest. It can be seen that this application further improves the accuracy of the prediction system for predicting the icing thickness of overhead transmission lines.

[0083] The method for predicting the icing thickness of an overhead transmission line provided by the embodiment of the present application includes obtaining historical data and future meteorological data of the overhead line to be predicted, where the historical data includes historical meteorological data, line data, equipment data, and historical icing thickness data; preprocessing the historical data to obtain preprocessed historical data, and performing feature engineering on the preprocessed historical data to obtain optimized historical data; dividing the optimized historical data into a training set and a test set, where the training set contains multiple validation sets, training multiple weak learner candidate models through the training set, calculating the scores of each weak learner candidate model on the validation set, and determining multiple strong learner models according to the scores of each weak learner candidate model; merging the optimized historical data and the future meteorological data into a data set, and dividing the data set into a total validation set and a total test set, training a meta-learner model through each strong learner model and the total validation set, and inputting the prediction results of each strong learner model on the total test set into the meta-learner model to obtain the final predicted icing thickness value of the total test set. This application predicts the future short-term icing thickness time series through multi-model fusion. By fusing and integrating multiple weak learners, the learning ability and generalization ability for icing thickness data are enhanced. At the same time, short-term meteorological data in the future prediction period is introduced. By comprehensively considering the covariate factors that affect the icing thickness evolution in the history and the future, the ability of the system to capture the temporal characteristics of the icing evolution is improved, and the accuracy of predicting the icing thickness of overhead transmission lines is improved.

[0084] Embodiment 2

[0085] As Figure 5 shown, it is a schematic structural diagram of an overhead transmission line icing thickness prediction device 500 in the embodiment of the present application. The device includes:

[0086] An acquisition module 510, configured to acquire historical data and future meteorological data of the overhead line to be predicted, where the historical data includes historical meteorological data, line data, equipment data, and historical icing thickness data;

[0087] A processing module 520, configured to preprocess the historical data to obtain preprocessed historical data, and perform feature engineering on the preprocessed historical data to obtain optimized historical data;

[0088] A training module 530, configured to divide the optimized historical data into a training set and a test set, where the training set includes multiple validation sets, train multiple weak learner candidate models through the training set, calculate the scores of each weak learner candidate model on the validation set, and determine multiple strong learner models according to the scores of each weak learner candidate model;

[0089] A prediction module 540, configured to merge the optimized historical data and the future meteorological data into a data set, divide the data set into a total validation set and a total test set, train a meta-learner model through each strong learner model and the total validation set, and input the prediction results of each strong learner model on the total test set into the meta-learner model to obtain the final predicted value of the icing thickness of the total test set.

[0090] The overhead transmission line icing thickness prediction device provided by the embodiment of the present application can implement each process of the overhead transmission line icing thickness prediction method corresponding to Embodiment 1, and can achieve the same technical effect. To avoid repetition, it will not be elaborated here.

[0091] The overhead transmission line icing thickness prediction device provided by the embodiment of the present application predicts the future short-term icing thickness time series through multi-model fusion. By fusing and integrating multiple weak learners, the learning ability and generalization ability of the icing thickness data are enhanced. At the same time, short-term meteorological data in the future prediction period is introduced. By comprehensively considering the covariate factors that affect the icing thickness evolution in history and the future, the ability of the system to capture the temporal characteristics of the icing evolution is improved, and the accuracy of the overhead transmission line icing thickness prediction is improved.

[0092] Embodiment 3

[0093] The embodiment of the present application also provides a computer device. Specifically, please refer to Figure 6 , Figure 6 which is the basic structural block diagram of the computer device in this embodiment.

[0094] The computer device 6 includes a memory 61, a processor 62, and a network interface 63 that are communicatively connected to each other via a system bus. It should be noted that only the computer device 6 with the memory 61, the processor 62, and the network interface 63 is shown in the figure. However, it should be understood that it is not required to implement all the shown components, and more or fewer components can be implemented alternatively. Among them, those skilled in the art of the present technology can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0095] The computer device can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The computer device can perform human-computer interaction with the user through a keyboard, a mouse, a remote control, a touchpad, a voice control device, etc.

[0096] The memory 61 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (such as SD or D-slot compatibility test memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disc, etc. In some embodiments, the memory 61 can be an internal storage unit of the computer device 6, such as the hard disk or memory of the computer device 6. In other embodiments, the memory 61 can also be an external storage device of the computer device 6, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device 6. Of course, the memory 61 can also include both the internal storage unit of the computer device 6 and its external storage device. In this embodiment, the memory 61 is generally used to store the operating system and various application software installed on the computer device 6, such as computer-readable instructions for the slot compatibility test method. In addition, the memory 61 can also be used to temporarily store various data that have been output or will be output.

[0097] In some embodiments, the processor 62 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other chips for predicting the ice thickness on overhead transmission lines. The processor 62 is generally used to control the overall operation of the computer device 6. In this embodiment, the processor 62 is used to run the computer-readable instructions stored in the memory 61 or process data, such as running the computer-readable instructions of the slot compatibility test method.

[0098] The network interface 63 may include a wireless network interface or a wired network interface. The network interface 63 is generally used to establish a communication connection between the computer device 6 and other electronic devices.

[0099] The computer device provided in this embodiment can execute the above-mentioned method for predicting the ice thickness on overhead transmission lines. Here, the method for predicting the ice thickness on overhead transmission lines can be the method for predicting the ice thickness on overhead transmission lines in each of the above embodiments.

[0100] Embodiment 4

[0101] This embodiment also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for predicting the ice thickness on overhead transmission lines in the embodiment are implemented.

[0102] In this embodiment, the computer-readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (such as SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the computer-readable storage medium may be an internal storage unit of the computer device, such as the hard disk or memory of the computer device. In other embodiments, the computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk equipped on the computer device, a smart media card (abbreviated as SMC), a secure digital (abbreviated as SD) card, a flash card, etc. Of course, the computer-readable storage medium may also include both the internal storage unit and the external storage device of the computer device. In this embodiment, the computer-readable storage medium is generally used to store the operating system and various application software installed on the computer device. In addition, the computer-readable storage medium can also be used to temporarily store various data that have been output or will be output.

[0103] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely illustrative. For example, the flowcharts and structure diagrams in the accompanying drawings show the possible architectures, functions, and operations of the devices, methods, and computer program products according to multiple embodiments of the present invention. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the structure diagram and / or flowchart, as well as the combination of blocks in the structure diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0104] In addition, in each embodiment of the present invention, each functional module or unit can be integrated together to form an independent part, or each module can exist separately, or two or more modules can be integrated to form an independent part.

[0105] If the above functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a smart phone, a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium can be a non-volatile storage medium or a volatile storage medium. For example, the storage medium can be: a USB flash drive, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk, or an optical disc, etc., which can store program codes.

[0106] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention.

Claims

1. A method for predicting the icing thickness of an overhead transmission line, characterized in that, The method includes: Obtaining historical data and future meteorological data of the overhead line to be predicted, where the historical data includes historical meteorological data, line data, equipment data, and historical ice thickness data; Preprocessing the historical data to obtain preprocessed historical data, and performing feature engineering on the preprocessed historical data to obtain optimized historical data; Dividing the optimized historical data into a training set and a test set, where the training set contains multiple validation sets, training multiple weak learner candidate models through the training set, calculating the scores of each weak learner candidate model on the validation set, and determining multiple strong learner models according to the scores of each weak learner candidate model; Merging the optimized historical data and the future meteorological data into a data set, and dividing the data set into a total validation set and a total test set, training a meta-learner model through each strong learner model and the total validation set, and inputting the prediction results of each strong learner model on the total test set into the meta-learner model to obtain the final ice thickness prediction value of the total test set.

2. The method for predicting the ice thickness of an overhead transmission line according to claim 1, wherein The preprocessing of the historical data to obtain preprocessed historical data includes: Sampling and aligning the historical data at a preset time interval to obtain aligned historical data; Determining the missing time period data in the aligned historical data, and filling the missing time period data with the mean value to obtain filled historical data; Determining the abnormal data in the filled historical data, and removing the abnormal data to obtain removed historical data; Determining the missing historical meteorological data in the removed historical data, and filling the missing historical meteorological data with Nan values to obtain the preprocessed historical data.

3. The method for predicting the ice thickness of an overhead transmission line according to claim 1, wherein, The feature engineering of the preprocessed historical data to obtain optimized historical data includes: Normalizing the historical meteorological data and the historical ice thickness data to obtain normalized historical data; Performing one-hot encoding on the line data and the equipment data to obtain one-hot encoded historical data; Performing principal component analysis on the one-hot encoded historical data, deleting the noise field columns in the one-hot encoded historical data, and performing data dimensionality reduction on the one-hot encoded historical data to obtain the optimized historical data.

4. The method for predicting the ice thickness of an overhead transmission line according to claim 1, characterized in that The dividing of the optimized historical data into a training set and a test set, where the training set contains multiple validation sets, includes: Dividing the optimized historical data into the training set and the test set according to a preset ratio; Using the Bagging algorithm to evenly divide the training set into multiple groups of first training sets by K-fold cross-validation; Dividing each first training set into a validation set and multiple second training sets.

5. The method for predicting the ice thickness of an overhead transmission line according to claim 4, wherein The calculating of the scores of each weak learner candidate model on the validation set and the determining of multiple strong learner models according to the scores of each weak learner candidate model includes: Predicting the validation set through each weak learner candidate model to obtain the ice thickness prediction values of each weak learner candidate model on the validation set; Obtain the actual icing thickness value of the verification set, and obtain the number of icing lines and the number of time series segments of the data in the verification set; According to the predicted icing thickness value, predicted icing thickness value, number of icing lines and number of time series segments of the verification set, and use the preset model score calculation formula to calculate the scores of each candidate weak learner model; Arrange the scores of each candidate weak learner model in ascending order, and determine the first preset number of candidate weak learner models as multiple strong learner models.

6. The method for predicting the ice thickness of an overhead transmission line according to claim 5, wherein, The preset model score calculation formula is: Where MAPE is the score of the weak learner candidate model, R is the number of ice-covered lines in the data of the validation set, T is the number of time series segments of the data in the validation set, d rt is the actual ice thickness value of the r-th line at the t-th time period in the validation set, and p rt is the predicted ice thickness value of the r-th line at the t-th time period in the validation set.

7. The method for predicting the ice coating thickness of an overhead transmission line according to claim 1, wherein Training the meta-learner model by each of the strong learner models and the total verification set includes: Predict the total verification set through each of the strong learner models to obtain the predicted icing thickness values of each strong learner model on the total verification set; Obtain the actual icing thickness value of the total verification set, use the predicted icing thickness value of the total verification set as the feature column, and use the actual icing thickness value of the total verification set as the target column to form a total training set, and train the meta-learner model through the total training set.

8. An icing thickness prediction device for an overhead transmission line, characterized in that, The device includes: An acquisition module, configured to acquire historical data and future meteorological data of an overhead line to be predicted, where the historical data includes historical meteorological data, line data, equipment data, and historical icing thickness data; A processing module, configured to preprocess the historical data to obtain preprocessed historical data, and perform feature engineering on the preprocessed historical data to obtain optimized historical data; A training module, configured to divide the optimized historical data into a training set and a test set, where the training set includes multiple verification sets, train multiple candidate weak learner models through the training set, calculate the scores of each candidate weak learner model on the verification set, and determine multiple strong learner models according to the scores of each candidate weak learner model; A prediction module, configured to merge the optimized historical data and the future meteorological data into a data set, divide the data set into a total verification set and a total test set, train a meta-learner model through each of the strong learner models and the total verification set, and input the prediction results of each strong learner model on the total test set into the meta-learner model to obtain the final predicted icing thickness value of the total test set.

9. A computer device, characterized in that, It includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, it implements the steps of the overhead transmission line icing thickness prediction method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements the steps of the overhead transmission line icing thickness prediction method according to any one of claims 1-7.

Images