Icing thickness prediction model construction method and icing thickness prediction method
Through the screening of multivariate linear regression model and target historical climate characteristic data, combined with historical ice covering data, an ice covering thickness prediction model was constructed, which solved the problem of low efficiency in ice covering thickness prediction in mountainous environments and achieved high efficiency under limited resource conditions , accurate prediction.
Patent Information
- Application Number
- CN202411278318.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-12
- Publication Date
- 2024-12-13
AI Technical Summary
In mountainous environments, the prediction efficiency of ice thickness is low, and the existing technology requires high data resources and computing resources, making it difficult to achieve accurate prediction.
A multivariate linear regression model is used to combine historical ice-covering data and climate characteristic data to build an ice-covering thickness prediction model. The model parameters are adjusted by screening target historical climate characteristic data and stepwise regression method to improve prediction accuracy and efficiency.
Under the conditions of limited data and computing resources, the rapid and accurate prediction of the ice cover thickness of the power facility improves the efficiency and accuracy of the ice cover thickness prediction.
Smart Images
Figure CN119129836A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of electric power technology, and in particular to a method, device, computer equipment, computer-readable storage medium, computer program product for constructing an ice thickness prediction model, and a method, device, computer equipment, computer-readable storage medium, and computer program product for predicting ice thickness. Background Art
[0002] With the continuous development of the power technology field, many remote areas have gradually achieved full coverage of electricity. However, some power facilities deployed in mountainous areas often have wires covered by ice in bad weather conditions, also known as icing. Severe icing will affect the normal operation and maintenance of power lines, so how to accurately predict the icing of power facilities has become one of the problems to be solved.
[0003] In traditional technologies, ice thickness is usually predicted based on physical models, machine learning models and other methods, such as finite element analysis, computational fluid dynamics, support vector machines, random forest models and other models. These models can handle some complex nonlinear relationships and achieve higher prediction accuracy in the field of ice prediction.
[0004] However, machine learning models have high requirements for training data and require high-quality input data. The processing process consumes a lot of resources. The construction and calculation process of the physical model also require strict parameter settings and a high-performance computer platform, which is not easy to achieve in mountainous environments where both data and computing resources are limited, resulting in inefficient ice thickness prediction. Summary of the invention
[0005] Based on this, it is necessary to provide a method, device, computer equipment, computer-readable storage medium and computer program product for constructing an ice thickness prediction model that can improve the efficiency of ice thickness prediction, as well as an ice thickness prediction method, device, computer equipment, computer-readable storage medium and computer program product in response to the above-mentioned technical problems.
[0006] In a first aspect, the present application provides a method for constructing an ice thickness prediction model, comprising:
[0007] Obtain historical icing data and historical climate characteristics data for power facilities in the target area;
[0008] Based on the data collection time of the historical climate characteristic data, filtering out target historical climate characteristic data from the historical climate characteristic data;
[0009] Based on the target historical climate characteristic data and the multivariate linear regression model, an initial ice thickness prediction model is constructed;
[0010] Based on the target historical climate characteristic data and historical ice cover data, the initial ice cover thickness prediction model is fitted, the regression coefficients of multiple independent variables in the initial ice cover thickness prediction model are determined, and the ice cover thickness prediction model is obtained.
[0011] In one embodiment, constructing an initial ice thickness prediction model based on the target historical climate characteristic data and a multivariate linear regression model includes:
[0012] Based on the correlation between the target historical climate characteristic data in different dimensions and the historical ice cover data, the target historical climate characteristic data is divided into a main independent variable and a secondary independent variable;
[0013] An initial ice thickness prediction model is constructed based on the main independent variables, the secondary independent variables and the multivariate linear regression model.
[0014] In one embodiment, constructing an initial ice thickness prediction model based on the target historical climate characteristic data and a multivariate linear regression model includes:
[0015] Determine an independent variable affecting ice thickness based on the target historical climate characteristic data and the interaction between the target historical climate characteristic data of different dimensions;
[0016] Based on the independent variables affecting ice thickness and the multivariate linear regression model, an initial ice thickness prediction model is constructed.
[0017] In one embodiment, after obtaining the ice thickness prediction model, the method further includes:
[0018] Evaluating the performance of the ice thickness prediction model to obtain a performance evaluation result;
[0019] Based on the performance evaluation result, the model parameters in the ice thickness prediction model are adjusted to obtain a target ice thickness prediction model.
[0020] In one embodiment, evaluating the performance of the ice thickness prediction model to obtain a performance evaluation result includes:
[0021] Determining a coefficient of determination of the ice thickness prediction model, wherein the coefficient of determination is used to evaluate the goodness of fit of the initial ice thickness prediction model;
[0022] Determine the significance test index value of each regression coefficient in the ice thickness prediction model, wherein the significance test index value is used to characterize the influence degree of each independent variable on the ice thickness;
[0023] Determining a mean square error of the ice thickness prediction model;
[0024] A performance evaluation result is determined based on at least one of the determination coefficient, the significance test index value, and the mean square error.
[0025] In one embodiment, adjusting the model parameters in the ice thickness prediction model based on the performance evaluation result to obtain a target ice thickness prediction model includes:
[0026] According to the performance evaluation result, a target independent variable having an influence degree greater than a preset influence degree threshold is screened out from the independent variables of the ice thickness prediction model by a stepwise regression method;
[0027] According to the selected target independent variable, the model parameters in the ice thickness prediction model are adjusted to obtain a target ice thickness prediction model.
[0028] In one of the embodiments, the performance evaluation result includes the prediction accuracy of the ice thickness prediction model;
[0029] The step of adjusting the model parameters in the ice thickness prediction model based on the performance evaluation result to obtain a target ice thickness prediction model includes:
[0030] When the prediction accuracy of the ice thickness prediction model is less than a preset prediction accuracy threshold, performing nonlinear transformation on the independent variable of the ice thickness prediction model to obtain an adjusted ice thickness prediction model;
[0031] Based on the target historical climate characteristic data and historical ice cover data, the adjusted ice cover thickness prediction model is refitted to obtain a target ice cover thickness prediction model.
[0032] In a second aspect, the present application also provides an ice thickness prediction method, comprising:
[0033] Obtain climate characteristic data of the target area within a preset time period;
[0034] Taking the climate characteristic data as input, calling a preset ice thickness prediction model to obtain the ice thickness of the power facilities in the target area;
[0035] The preset ice thickness prediction model is constructed based on the steps of any embodiment of the ice thickness prediction model construction method.
[0036] In a third aspect, the present application also provides an ice thickness prediction model construction device, comprising:
[0037] A historical data acquisition module, used to acquire historical ice coverage data and historical climate characteristic data of power facilities in the target area;
[0038] A data screening module, for screening target historical climate characteristic data from the historical climate characteristic data based on the data collection time of the historical climate characteristic data;
[0039] A model building module, used to build an initial ice thickness prediction model based on the target historical climate characteristic data and a multivariate linear regression model;
[0040] The model fitting module is used to fit the initial ice thickness prediction model based on the target historical climate characteristic data and historical ice cover data, determine the regression coefficients of multiple independent variables in the initial ice thickness prediction model, and obtain the ice thickness prediction model.
[0041] In a fourth aspect, the present application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any embodiment of the method for constructing an ice thickness prediction model when executing the computer program.
[0042] In a fifth aspect, the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any embodiment of a method for constructing an ice thickness prediction model.
[0043] In a sixth aspect, the present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of any embodiment of the method for constructing an ice thickness prediction model.
[0044] In a seventh aspect, the present application also provides an ice thickness prediction device, comprising:
[0045] A data acquisition module is used to acquire climate characteristic data of a target area within a preset time period;
[0046] An ice thickness prediction module is used to use the climate characteristic data as input, call a preset ice thickness prediction model, and obtain the ice thickness of the power facilities in the target area;
[0047] The preset ice thickness prediction model is constructed based on the steps of any embodiment of the ice thickness prediction model construction method.
[0048] In an eighth aspect, the present application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any embodiment of an ice thickness prediction method when executing the computer program.
[0049] In a ninth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of any embodiment of the ice thickness prediction method.
[0050] In a tenth aspect, the present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of any embodiment of the ice thickness prediction method.
[0051] The above ice thickness prediction method, device, computer equipment, computer readable storage medium and computer program product are different from traditional solutions. They combine historical ice data, historical climate characteristic data and multivariate linear regression models. Compared with physical models and machine learning models in traditional solutions, multivariate linear regression models consume relatively less computing resources and process data more quickly, so that the ice thickness of power facilities in the target area can be quickly predicted in a short time, thereby improving the prediction efficiency. Furthermore, based on the data collection time of historical climate characteristic data, the target historical climate characteristic data selected from the historical climate characteristic data can better reflect the future ice situation, and then based on the target historical climate characteristic data and the historical ice thickness data, the ice thickness prediction model obtained by fitting the initial ice thickness prediction model can well handle the correlation between climate and ice thickness, so that relatively accurate prediction results can still be provided under limited data conditions. Therefore, the above scheme can still efficiently predict the ice thickness of power equipment in mountainous environments where data resources and computing resources are relatively scarce. BRIEF DESCRIPTION OF THE DRAWINGS
[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.
[0053] FIG1 is a diagram of an application environment of an ice thickness prediction model construction method or an ice thickness prediction method in one embodiment;
[0054] Figure 2 A schematic diagram of a process for constructing an ice thickness prediction model in one embodiment;
[0055] Figure 3 A schematic diagram of a flow chart of a method for constructing an ice thickness prediction model in another embodiment;
[0056] Figure 4 A schematic flow chart of a performance evaluation step of a method for building an ice thickness prediction model in one embodiment;
[0057] Figure 5 A schematic diagram of a flow chart of a method for constructing an ice thickness prediction model in another embodiment;
[0058] Figure 6 A schematic diagram of a process for constructing an ice thickness prediction model in a detailed embodiment;
[0059] Figure 7 A schematic diagram of a flow chart of an ice thickness prediction method in one embodiment;
[0060] Figure 8 It is a structural block diagram of an ice thickness prediction model building device in one embodiment;
[0061] Fig. 9 is a structural block diagram of an ice thickness prediction device in one embodiment;
[0062] FIG. 10 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION
[0063] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0064] The ice thickness prediction model construction method and ice thickness prediction method provided in the embodiment of the present application can be applied to the application environment shown in Figure 1. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers.
[0065] When the ice thickness prediction model construction method is applied in the application environment shown in FIG1 , the operator uploads the historical ice data and historical climate characteristic data of the power facilities in the target area to the server 104 through the terminal 102, or the server 104 directly obtains the above data from the data storage system. Then, based on the data collection time of the historical climate characteristic data, the server 104 selects the target historical climate characteristic data from the historical climate characteristic data, constructs an initial ice thickness prediction model based on the target historical climate characteristic data and the multivariate linear regression model, fits the initial ice thickness prediction model based on the target historical climate characteristic data and the historical ice data, determines the regression coefficients of multiple independent variables in the initial ice thickness prediction model, and obtains the ice thickness prediction model.
[0066] When the ice thickness prediction method is applied in the application environment shown in FIG1 , the operator uploads the climate characteristic data of the target area within the preset time period to the server 104 through the terminal 102. The terminal 102 may be a terminal of a weather station or other weather monitoring base, or the server 104 may directly obtain the above data from the data storage system. Then, the server 104 calls a preset ice thickness prediction model, which may be pre-stored in the data storage system of the server 104 and constructed by the steps in the above-mentioned ice thickness prediction model construction method embodiment. Further, the ice thickness of the power facilities in the target area is obtained by processing the climate characteristic data through the ice thickness prediction model.
[0067] The terminal 102 may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart car devices, projection devices, etc. Portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. Head-mounted devices may be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. The server 104 may be an independent physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides cloud computing services.
[0068] In an exemplary embodiment, Figure 2 As shown, a method for building an ice thickness prediction model is provided, and the method is applied to the server 104 in FIG. 1 as an example for description, including the following steps:
[0069] S100, obtaining historical ice coverage data and historical climate characteristic data of power facilities in a target area.
[0070] The target area refers to the geographical area where ice thickness detection is required, including but not limited to mountainous areas and other areas with complex terrain. Power facilities include but are not limited to transmission lines, substations, distribution lines and other facilities exposed to the natural environment. These power facilities may be affected by ice under extreme climatic conditions. Historical ice data includes but is not limited to the thickness and duration of ice formed on the power facilities in the target area during a specific period of time in the past. Historical climate characteristic data includes but is not limited to meteorological parameters closely related to ice formation, such as temperature, humidity, precipitation, and wind speed.
[0071] For example, the staff can obtain the historical icing data and historical climate characteristic data of the power facilities in the target area by retrieving relevant historical databases or weather station records. Specifically, the staff can collect historical climate characteristic data through the Meteorological Bureau or through meteorological monitoring equipment installed near the power facilities, and obtain the historical icing data of the power facilities through on-site monitoring equipment near the power facilities or historical inspection reports.
[0072] S200, based on the data collection time of the historical climate characteristic data, filter out target historical climate characteristic data from the historical climate characteristic data.
[0073] The data collection time refers to the timestamp recorded when the historical climate characteristic data is obtained, usually expressed as a specific date and time. The target historical climate characteristic data refers to some climate characteristic data that meet specific conditions and are used to build an ice thickness prediction model, such as historical climate characteristic data that has a greater impact on the ice thickness of power facilities.
[0074] Specifically, the acquired historical climate characteristic data are screened based on the data collection time. For example, the historical climate characteristic data in the previous month is more likely to predict the ice thickness of the power facilities in the target area in the next week than the historical climate characteristic data a year ago. In addition, the data can also be screened based on the periodic characteristics of the climate phenomenon. For example, the ice coverage in the target area usually occurs between November and December. When screening the data, only the historical climate characteristic data between November and December of previous years are retained.
[0075] S300, based on the target historical climate characteristic data and the multivariate linear regression model, constructs an initial ice thickness prediction model.
[0076] Among them, the multivariate linear regression model is used to study the influence relationship of multiple independent variables on one dependent variable. In this embodiment, the independent variables are climate characteristic data of different dimensions, and the dependent variable is ice thickness.
[0077] Specifically, the multivariate linear regression model used in this embodiment can be expressed by formula (1):
[0078] (1)
[0079] in, is the intercept, , , is the regression coefficient of each variable, which represents the influence of the independent variable on the dependent variable. Since the multivariate linear regression model may not perfectly fit the actual historical data during the fitting process, an error term is added. In formula (1), The error term is the preset error term. Taking the selected target historical climate characteristic data including temperature, humidity and precipitation in the past month as an example, these factors are input into the multivariate linear regression model as independent variables to construct the initial ice thickness prediction model, including three independent variables: , , , representing temperature, humidity and precipitation respectively.
[0080] S400, based on the target historical climate characteristic data and the historical ice cover data, fit the initial ice cover thickness prediction model, determine the regression coefficients of multiple independent variables in the initial ice cover thickness prediction model, and obtain the ice cover thickness prediction model.
[0081] Following the above steps, the constructed initial ice thickness prediction model is fitted, for example, by using the least squares method, and the regression coefficients of multiple independent variables in the initial ice thickness prediction model are determined according to the target historical climate characteristic data and historical ice cover data. This process is repeated through iterations, and the regression coefficients are adjusted multiple times. Finally, the initial ice thickness prediction model tends to be stable and can reliably predict the ice thickness of power facilities in the target area under future climate conditions.
[0082] It should be noted that the ice thickness prediction model in this embodiment has good model interpretability. The ice thickness prediction model provides a clear relationship between the independent variables (climate characteristic data of different dimensions) and the dependent variable (ice thickness), and has low data demand during the construction process. It is especially suitable for mountainous environments with scarce data and can provide effective predictions under limited data conditions.
[0083] The above ice thickness prediction method is different from the traditional solution. It combines historical ice data, historical climate characteristic data and multivariate linear regression model. Compared with the physical model and machine learning model in the traditional solution, the multivariate linear regression model consumes relatively less computing resources and processes data more quickly, so that the ice thickness of power facilities in the target area can be quickly predicted in a short time, improving the prediction efficiency. Furthermore, based on the data collection time of the historical climate characteristic data, the target historical climate characteristic data selected from the historical climate characteristic data can better reflect the future ice situation, and then based on the target historical climate characteristic data and the historical ice thickness data, the ice thickness prediction model obtained by fitting the initial ice thickness prediction model can well handle the correlation between climate and ice thickness, so that it can still provide relatively accurate prediction results under limited data conditions. Therefore, the above scheme can still efficiently predict the ice thickness of power equipment in mountainous environments where data resources and computing resources are relatively scarce.
[0084] Following the above embodiment, not all dimensional climate characteristic data have a significant impact on ice thickness, so before constructing the initial ice thickness prediction model, the target historical climate characteristic data of different dimensions can be divided into primary independent variables and secondary independent variables. In an exemplary embodiment, S300 includes: based on the degree of correlation between the target historical climate characteristic data of different dimensions and the historical ice data, the target historical climate characteristic data is divided into primary independent variables and secondary independent variables, and the initial ice thickness prediction model is constructed according to the primary independent variables, the secondary independent variables and the multivariate linear regression model.
[0085] The degree of correlation refers to the strength of the influence between two variables, which can be expressed by the correlation coefficient. For example, the value range of the correlation coefficient is between -1 and 1. A value close to 1 or -1 indicates a strong correlation, and a value close to 0 indicates no correlation. The main independent variable refers to the independent variable that has a significant impact on the dependent variable in the regression analysis, and the secondary independent variable refers to the independent variable that has a smaller impact on the dependent variable.
[0086] Specifically, before constructing the initial ice thickness prediction model, the target historical climate characteristic data of different dimensions are first analyzed to determine the degree of correlation between the target historical climate characteristic data of different dimensions and the historical ice cover data, wherein the degree of correlation can be quantified by calculating the correlation coefficient, and then determine which dimensions of climate characteristic data will be used as the main independent variable or the secondary independent variable. Exemplarily, for climate characteristic data of different dimensions (such as temperature, humidity, precipitation, wind speed, etc.), the correlation coefficient between them and ice cover thickness is calculated. If the correlation coefficient of climate characteristic data of a certain dimension is high (such as an absolute value greater than 0.7), it is considered that the climate characteristic data of this dimension has a significant impact on the ice cover thickness and is marked as the main independent variable; while climate characteristic data with a lower correlation coefficient (such as an absolute value less than 0.3) is marked as a secondary independent variable.
[0087] After determining the main independent variables and secondary independent variables, the main independent variables have a greater correlation with the ice thickness and can be directly used to construct the initial ice thickness prediction model. The secondary independent variables have a lower correlation with the ice thickness, but in some cases they can also play an auxiliary role in the overall accuracy of the initial ice thickness prediction model. Therefore, in practical applications, you can choose to use only the main independent variables to simplify the initial ice thickness prediction model, or you can introduce the secondary independent variables as correction items into the initial ice thickness prediction model to improve the overall accuracy of the initial ice thickness prediction model.
[0088] In this embodiment, by distinguishing between the main independent variables and the secondary independent variables, the structure of the initial ice thickness prediction model can be simplified, and irrelevant or weakly correlated independent variables introduced into the initial ice thickness prediction model can be reduced, thereby reducing the risk of overfitting. At the same time, retaining the secondary independent variables as correction items can improve the overall accuracy of the initial ice thickness prediction model, thereby helping power facility managers to more accurately estimate ice risks.
[0089] In addition to distinguishing between primary and secondary independent variables, climate characteristic data of different dimensions may not affect ice thickness alone, so it is also necessary to consider the impact of the interaction between climate characteristic data of different dimensions on ice thickness. In an exemplary embodiment, S300 includes: determining independent variables affecting ice thickness based on target historical climate characteristic data and the interaction between target historical climate characteristic data of different dimensions, and constructing an initial ice thickness prediction model based on the independent variables affecting ice thickness and a multiple linear regression model.
[0090] The interaction refers to the mutual influence between two or more independent variables, which may cause the total effect of the two or more independent variables on the dependent variable to be different from the sum of the effects of each independent variable acting alone. Therefore, in this embodiment, it is not necessary to use the target historical climate characteristic data of each dimension as an independent variable alone, but the interaction terms of the target historical climate characteristic data of different dimensions may be used as independent variables.
[0091] For example, temperature and humidity may have an impact on ice thickness individually, but the combined effect of the two may have a more significant impact on ice thickness. When the temperature is low and the humidity is high, ice may be more likely to form. Therefore, an interaction term between the two can be constructed to quantify this effect, such as multiplying temperature and humidity to obtain a new variable (temperature*humidity), and using this new variable as the independent variable affecting ice thickness in the initial ice thickness prediction model.
[0092] In this embodiment, considering that under extreme climatic conditions, the interaction of climate characteristic data of different dimensions often leads to nonlinear results, a simple linear model may not be able to accurately predict the ice thickness in this case. Therefore, this embodiment introduces interaction terms of climate characteristic data of different dimensions to construct an initial ice thickness prediction model that can capture the impact of more complex climatic conditions on ice thickness, thereby improving the prediction accuracy.
[0093] Following the above embodiment, after obtaining the ice thickness prediction model, the ice thickness prediction model can be further evaluated and optimized to improve the prediction accuracy. Figure 3 As shown, after obtaining the ice thickness prediction model, the method further includes:
[0094] S500: Evaluate the performance of the ice thickness prediction model to obtain a performance evaluation result.
[0095] S600: Based on the performance evaluation result, adjust the model parameters in the ice thickness prediction model to obtain a target ice thickness prediction model.
[0096] Among them, evaluating the performance of the ice thickness prediction model refers to testing and analyzing the prediction ability of the ice thickness prediction model through certain indicators and methods to determine the accuracy and stability of the ice thickness prediction model. For example, the performance evaluation indicators used to evaluate the ice thickness prediction model may include but are not limited to mean square error, mean absolute error, determination coefficient, residual analysis, etc., which are used to quantify the prediction error and fitting effect of the ice thickness prediction model.
[0097] For example, in order to evaluate the generalization ability of the model, the data set can be divided into a training set and a test set by cross-validation, such as the K-fold cross-validation method, wherein the training set is used to build the model and the test set is used to evaluate the model performance. Using the test set, the values of performance evaluation indicators such as the mean square error, mean absolute error, and determination coefficient of the ice thickness prediction model are calculated to evaluate the performance of the ice thickness prediction model. For example, a lower mean square error and mean absolute error, as well as a higher determination coefficient, indicate that the ice thickness prediction model has good prediction ability.
[0098] In this embodiment, by evaluating the performance of the ice thickness prediction model, the quality of the current ice thickness prediction model can be judged, and the obtained evaluation results can provide data support for the subsequent adjustment of the model parameters of the ice thickness prediction model, so that the adjusted model has higher prediction accuracy.
[0099] Specifically, the above performance evaluation indicators include but are not limited to the coefficient of determination, the significance test index value, and the mean square error. In an exemplary embodiment, Figure 4 As shown, S500 includes:
[0100] S510, determining a determination coefficient of an ice thickness prediction model, where the determination coefficient is used to evaluate the goodness of fit of the initial ice thickness prediction model.
[0101] S520, determining the significance test index value of each regression coefficient in the ice thickness prediction model, where the significance test index value is used to characterize the influence of each variable on the ice thickness.
[0102] S530, determining a mean square error of the ice thickness prediction model.
[0103] S540, determining a performance evaluation result based on at least one of the determination coefficient, the significance test index value, and the mean square error.
[0104] Among them, the coefficient of determination (R²) can be used to measure the goodness of fit of the regression model to the data. The range of the coefficient of determination is 0 to 1. The closer it is to 1, the better the model fit. The significance test index value is used to measure whether the influence of each variable on the dependent variable is significant. The significance test method includes but is not limited to t-test and p-value analysis. The mean square error represents the average value of the sum of squares of the errors between the model prediction value and the actual value. The smaller the mean square error, the higher the prediction accuracy of the model.
[0105] Specifically, the coefficient of determination indicates the proportion of the variance of the dependent variable explained by the independent variable to the total variance. If the coefficient of determination of the ice thickness prediction model is close to 1, it means that the ice thickness prediction model well explains the change of ice thickness with the independent variable and has good fitting performance. However, an excessively high coefficient of determination may also indicate that the ice thickness prediction model has the risk of overfitting. At this time, other performance evaluation indicators can be combined to comprehensively evaluate the model performance. Furthermore, in the ice thickness prediction model, each independent variable has a corresponding regression coefficient, which is used to describe the influence of the independent variable on the dependent variable (ice thickness). Taking the p-value analysis to evaluate whether the regression coefficient of each variable is significant as an example, if the p-value of the regression coefficient of a certain independent variable is lower than the preset coefficient threshold (such as 0.05), it is considered that the independent variable has a significant effect on the ice thickness. On the contrary, if the p-value is very high, it indicates that the independent variable has little effect on the ice thickness. The mean square error measures the average deviation of the model prediction results, that is, the average square error between the model prediction value and the actual value. The smaller the mean square error, the higher the accuracy of the ice thickness prediction model in predicting ice thickness. Based on at least one of the above-mentioned determination coefficient, significance test index value, and mean square error, the performance of the ice thickness prediction model is evaluated to determine the performance evaluation result.
[0106] In this embodiment, the overall performance of the ice thickness prediction model is evaluated by at least one of the determination coefficient, the significance test index value, and the mean square error. If the value of one of the performance evaluation indicators is not ideal, it can help to subsequently adjust the model structure or model parameters of the ice thickness prediction model in a targeted manner, thereby improving the prediction accuracy of the ice thickness prediction model.
[0107] Furthermore, after obtaining the evaluation result, the model structure or model parameters of the ice thickness prediction model can be adjusted in a targeted manner. In an exemplary embodiment, S600 includes:
[0108] According to the performance evaluation results, the stepwise regression method is used to screen out the target independent variables whose influence degree is greater than the preset influence degree threshold from the independent variables of the ice thickness prediction model. According to the screened target independent variables, the model parameters in the ice thickness prediction model are adjusted to obtain the target ice thickness prediction model.
[0109] Among them, the stepwise regression method is a statistical method for selecting independent variables. It constructs the optimal regression model by gradually introducing or eliminating independent variables and evaluating their impact on the dependent variables. The stepwise regression method includes forward selection, backward elimination and two-way stepwise regression.
[0110] For example, in the stepwise regression process, the independent variables that have a significant impact on the ice thickness prediction are specifically screened out according to the influence of each independent variable in the performance evaluation result on the ice thickness prediction. In this process, it can be evaluated whether each independent variable in the ice thickness prediction model exceeds a preset influence degree threshold, and the influence degree can be characterized by the significance test index value in the above embodiment.
[0111] After screening out the target independent variables that have a significant impact on the prediction of ice thickness, the model parameters of the ice thickness prediction model need to be readjusted. Specifically, by eliminating insignificant independent variables, the ice thickness prediction model becomes more concise, and the target ice thickness prediction model is obtained. Specifically, in the forward selection algorithm, one of the most significant independent variables is selected to be added to the model each time until there are no significant variables to be added. In the backward elimination method, starting from the ice thickness prediction model containing all independent variables, the least significant independent variable is eliminated each time until all remaining independent variables are significant. The two-way stepwise regression combines the characteristics of forward selection and backward elimination to comprehensively screen the independent variables. In addition, after eliminating insignificant independent variables, the initial ice prediction model can be reconstructed, and the initial ice prediction model can be fitted again to obtain the target ice thickness prediction model.
[0112] In this embodiment, the stepwise regression method is used to screen out the target independent variables that have a significant impact on the ice thickness prediction, and the model parameters of the ice thickness prediction model are adjusted, which can significantly improve the accuracy and stability of the ice thickness prediction model. The final target ice thickness prediction model not only has higher prediction capabilities, but also can better adapt to the complex climatic conditions in the actual operation of power facilities, reducing the risk of erroneous predictions.
[0113] In addition to eliminating independent variables that do not significantly affect the ice thickness prediction, the form of the independent variables can also be adjusted, such as nonlinear transformation, to improve the prediction accuracy of the model. In an exemplary embodiment, the performance evaluation results include the prediction accuracy of the ice thickness prediction model, such as Figure 5 As shown, S600 includes:
[0114] S610: When the prediction accuracy of the ice thickness prediction model is less than a preset prediction accuracy threshold, a nonlinear transformation is performed on the independent variable of the ice thickness prediction model to obtain an adjusted ice thickness prediction model.
[0115] S620, based on the target historical climate characteristic data and the historical ice cover data, refit the adjusted ice cover thickness prediction model to obtain the target ice cover thickness prediction model.
[0116] Among them, the nonlinear transformation transforms the independent variable nonlinearly to capture the complex relationship between the independent variable and the dependent variable, for example, the independent variable is squared, squared or logarithmized. Specifically, when the prediction accuracy of the ice thickness prediction model is less than the preset prediction accuracy threshold, it indicates that the model structure of the current ice thickness prediction model fails to fully capture the relationship between the independent variable and the dependent variable. This is because the linear regression model assumes that the relationship between the independent variable and the dependent variable is linear, but in practical applications, the relationship between the independent variable and the dependent variable is often more complex. Therefore, the shortcomings of the linear regression model can be compensated by performing a nonlinear transformation on the independent variable. After the nonlinear transformation of the independent variable, the adjusted ice thickness prediction model is refitted. The fitting process can also be completed by using the least squares method or other fitting methods through Python, and the regression coefficients of each variable are re-determined to obtain the target ice thickness prediction model.
[0117] In this embodiment, by performing performance evaluation and nonlinear adjustment on the ice thickness prediction model, the complex relationship between the independent variables and the dependent variables can be better captured. The obtained target ice thickness prediction model has higher prediction accuracy and can better adapt to complex practical application scenarios.
[0118] In order to make a clearer description of the ice thickness prediction model construction method provided by the present application, the following is a detailed embodiment and attached Figure 6 The method of constructing the ice thickness prediction model is explained.
[0119] S601, obtaining historical ice coverage data and historical climate characteristic data of power facilities in a target area, and filtering out target historical climate characteristic data from the historical climate characteristic data based on the data collection time of the historical climate characteristic data.
[0120] S602, based on the target historical climate characteristic data and the interaction between the target historical climate characteristic data of different dimensions, determine the independent variables affecting the ice thickness, and construct an initial ice thickness prediction model based on the independent variables affecting the ice thickness and the multivariate linear regression model.
[0121] S603, based on the target historical climate characteristic data and historical ice cover data, fit the initial ice cover thickness prediction model, determine the regression coefficients of multiple independent variables in the initial ice cover thickness prediction model, and obtain the ice cover thickness prediction model.
[0122] S604, determining the determination coefficient of the ice thickness prediction model, where the determination coefficient is used to evaluate the goodness of fit of the initial ice thickness prediction model and determine the significance test index value of each regression coefficient in the ice thickness prediction model.
[0123] S605, determining a mean square error of the ice thickness prediction model, and determining a performance evaluation result based on at least one of a determination coefficient, a significance test index value, and the mean square error.
[0124] S606: When the prediction accuracy of the ice thickness prediction model is less than a preset prediction accuracy threshold, a nonlinear transformation is performed on the independent variable of the ice thickness prediction model to obtain an adjusted ice thickness prediction model.
[0125] S607, based on the target historical climate characteristic data and the historical ice cover data, refit the adjusted ice cover thickness prediction model to obtain the ice cover thickness prediction model.
[0126] In an exemplary embodiment, Figure 7 As shown, a method for predicting ice thickness is provided, which is described by taking the method applied to the server 104 in FIG. 1 as an example, and includes the following steps:
[0127] S700, obtaining climate characteristic data of a target area within a preset time period.
[0128] S800, using climate characteristic data as input, calling a preset ice thickness prediction model to obtain the ice thickness of power facilities in the target area.
[0129] Among them, the ice thickness prediction model is constructed by any of the steps in the above-mentioned ice thickness prediction model construction method embodiments, and the model construction process is not repeated here. The preset time period can be a certain period of time in the future, or it can be real-time. This is because the ice thickness prediction model constructed by the steps in the above-mentioned ice thickness prediction model construction method embodiment has a small data and computing resource requirement, and can quickly process and analyze climate characteristic data, thereby realizing real-time or near real-time ice thickness prediction.
[0130] For example, taking the preset time period of the next 48 hours as an example, the staff can obtain the climate characteristic data of the target area in the next 48 hours from the meteorological station, satellite remote sensing and other monitoring equipment, including but not limited to temperature, humidity, wind speed and precipitation, etc. Further, the server calls the pre-built ice thickness prediction model, predicts the ice thickness based on the climate characteristic data, and obtains the ice thickness of the power facilities in the target area.
[0131] In addition, the ice thickness prediction model can also be integrated into the real-time meteorological data system to realize dynamic prediction and warning functions. Specifically, by real-time monitoring of meteorological data and real-time ice thickness prediction through the ice thickness prediction model, when the ice thickness is greater than the preset ice thickness threshold, the warning mechanism is triggered in time, such as sounding an alarm and sending an alarm message to the staff, so as to take preventive measures in time to reduce the potential damage caused by ice to power facilities.
[0132] In this embodiment, by acquiring climate characteristic data within a preset time period and predicting the ice thickness through an ice thickness prediction model, the ice thickness can be obtained quickly and accurately. Different from the physical model and deep learning model in the traditional scheme, the ice thickness prediction model in the above scheme is constructed based on a multivariate linear regression model. Therefore, the data processing process during ice prediction is relatively simple, and no high-performance computing platform is required. It can also quickly process and analyze climate characteristic data, thereby improving the efficiency of ice prediction.
[0133] It should be understood that, although the steps in the flowcharts involved in the above embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.
[0134] Based on the same inventive concept, the embodiment of the present application also provides an ice thickness prediction model construction device for implementing the ice thickness prediction model construction method involved above, and an ice thickness prediction device for implementing the ice thickness prediction method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more ice thickness prediction model construction device embodiments provided below can refer to the limitations of the ice thickness prediction model construction method above, and the specific limitations in one or more ice thickness prediction device embodiments provided below can refer to the limitations of the ice thickness prediction method above, and will not be repeated here.
[0135] In an exemplary embodiment, Figure 8As shown, an ice thickness prediction model building device 800 is provided, comprising: a historical data acquisition module 810, a data screening module 820, a model building module 830 and a model fitting module 840, wherein:
[0136] The historical data acquisition module 810 is used to acquire the historical ice coverage data and historical climate characteristic data of the power facilities in the target area;
[0137] A data screening module 820 is used to screen out target historical climate characteristic data from the historical climate characteristic data based on the data collection time of the historical climate characteristic data;
[0138] A model building module 830 is used to build an initial ice thickness prediction model based on target historical climate characteristic data and a multivariate linear regression model;
[0139] The model fitting module 840 is used to fit the initial ice thickness prediction model based on the target historical climate characteristic data and the historical ice cover data, determine the regression coefficients of multiple independent variables in the initial ice thickness prediction model, and obtain the ice thickness prediction model.
[0140] In an exemplary embodiment, the model building module 830 is also used to divide the target historical climate characteristic data into primary independent variables and secondary independent variables based on the degree of correlation between the target historical climate characteristic data in different dimensions and the historical ice cover data, and to construct an initial ice cover thickness prediction model based on the primary independent variables, secondary independent variables and a multivariate linear regression model.
[0141] In an exemplary embodiment, the model building module 830 is also used to determine the independent variables affecting the ice thickness based on the target historical climate characteristic data and the interaction between the target historical climate characteristic data of different dimensions, and to construct an initial ice thickness prediction model based on the independent variables affecting the ice thickness and the multivariate linear regression model.
[0142] In an exemplary embodiment, the ice thickness prediction model building device 800 is also used to evaluate the performance of the ice thickness prediction model to obtain a performance evaluation result, and based on the performance evaluation result, adjust the model parameters in the ice thickness prediction model to obtain a target ice thickness prediction model.
[0143] In an exemplary embodiment, the ice thickness prediction model building device 800 is also used to determine the determination coefficient of the ice thickness prediction model, the determination coefficient is used to evaluate the goodness of fit of the initial ice thickness prediction model, determine the significance test index value of each regression coefficient in the ice thickness prediction model, the significance test index value is used to characterize the degree of influence of each variable on the ice thickness, determine the mean square error of the ice thickness prediction model, and determine the performance evaluation result based on at least one of the determination coefficient, the significance test index value, and the mean square error.
[0144] In an exemplary embodiment, the ice thickness prediction model building device 800 is also used to screen out target independent variables whose influence degree is greater than a preset influence degree threshold from the independent variables of the ice thickness prediction model according to the performance evaluation results through a stepwise regression method, and adjust the model parameters in the ice thickness prediction model according to the screened target independent variables to obtain the target ice thickness prediction model.
[0145] In an exemplary embodiment, the ice thickness prediction model building device 800 is also used to perform nonlinear transformation on the independent variables of the ice thickness prediction model when the prediction accuracy of the ice thickness prediction model is less than a preset prediction accuracy threshold to obtain an adjusted ice thickness prediction model, and re-fit the adjusted ice thickness prediction model based on the target historical climate characteristic data and historical ice cover data to obtain the target ice thickness prediction model.
[0146] In an exemplary embodiment, Fig. 9 As shown, an ice thickness prediction device 900 is provided, comprising: a data acquisition module 910 and an ice thickness prediction module 920, wherein:
[0147] The data acquisition module 910 is used to acquire the climate characteristic data of the target area within a preset time period;
[0148] The ice thickness prediction module 920 is used to use the climate characteristic data as input, call a preset ice thickness prediction model, and obtain the ice thickness of the power facilities in the target area;
[0149] Among them, the preset ice thickness prediction model is constructed based on the steps of any embodiment of the ice thickness prediction model construction method.
[0150] Each module in the ice thickness prediction model building device or ice thickness prediction device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.
[0151] In an exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be shown in FIG10. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data such as an ice thickness prediction model. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for constructing an ice thickness prediction model and an ice thickness prediction method are implemented.
[0152] Those skilled in the art will understand that the structure shown in FIG. 10 is merely a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.
[0153] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps of any embodiment of the method for constructing an ice thickness prediction model are implemented.
[0154] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of any embodiment of the method for constructing an ice thickness prediction model are implemented.
[0155] In one embodiment, a computer program product is provided, including a computer program, which, when executed by a processor, implements the steps of any embodiment of the method for constructing an ice thickness prediction model.
[0156] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps of any ice thickness prediction method embodiment are implemented.
[0157] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of any embodiment of the ice thickness prediction method are implemented.
[0158] In one embodiment, a computer program product is provided, comprising a computer program, which implements the steps of any ice thickness prediction method embodiment when executed by a processor.
[0159] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.
[0160] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.
[0161] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.
[0162] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.
Claims
1. A method for constructing an ice thickness prediction model. It is characterized in that The method comprises: Obtain historical icing data and historical climate characteristics data for power facilities in the target area; Based on the data collection time of the historical climate characteristic data, filtering out target historical climate characteristic data from the historical climate characteristic data; Based on the target historical climate characteristic data and the multivariate linear regression model, an initial ice thickness prediction model is constructed; Based on the target historical climate characteristic data and historical ice cover data, the initial ice cover thickness prediction model is fitted, the regression coefficients of multiple independent variables in the initial ice cover thickness prediction model are determined, and the ice cover thickness prediction model is obtained.
2. The method according to claim 1, It is characterized in that The initial ice thickness prediction model is constructed based on the target historical climate characteristic data and the multivariate linear regression model, including: Based on the correlation between the target historical climate characteristic data in different dimensions and the historical ice cover data, the target historical climate characteristic data is divided into a main independent variable and a secondary independent variable; An initial ice thickness prediction model is constructed based on the main independent variables, the secondary independent variables and the multivariate linear regression model.
3. The method according to claim 1, It is characterized in that The initial ice thickness prediction model is constructed based on the target historical climate characteristic data and the multivariate linear regression model, including: Determine an independent variable affecting ice thickness based on the target historical climate characteristic data and the interaction between the target historical climate characteristic data of different dimensions; Based on the independent variables affecting ice thickness and the multivariate linear regression model, an initial ice thickness prediction model is constructed.
4. The method according to any one of claims 1 to 3, It is characterized in that After obtaining the ice thickness prediction model, the method further includes: Evaluating the performance of the ice thickness prediction model to obtain a performance evaluation result; Based on the performance evaluation result, the model parameters in the ice thickness prediction model are adjusted to obtain a target ice thickness prediction model.
5. The method according to claim 4, It is characterized in that The step of evaluating the performance of the ice thickness prediction model to obtain a performance evaluation result includes: Determining a coefficient of determination of the ice thickness prediction model, wherein the coefficient of determination is used to evaluate the goodness of fit of the initial ice thickness prediction model; Determine the significance test index value of each regression coefficient in the ice thickness prediction model, wherein the significance test index value is used to characterize the influence degree of each independent variable on the ice thickness; Determining a mean square error of the ice thickness prediction model; A performance evaluation result is determined based on at least one of the determination coefficient, the significance test index value, and the mean square error.
6. The method according to claim 4, It is characterized in that The adjusting the model parameters in the ice thickness prediction model based on the performance evaluation result to obtain a target ice thickness prediction model includes: According to the performance evaluation result, a target independent variable having an influence degree greater than a preset influence degree threshold is screened out from the independent variables of the ice thickness prediction model by a stepwise regression method; According to the selected target independent variable, the model parameters in the ice thickness prediction model are adjusted to obtain a target ice thickness prediction model.
7. The method according to claim 4, It is characterized in that The performance evaluation result includes the prediction accuracy of the ice thickness prediction model; The adjusting the model parameters in the ice thickness prediction model based on the performance evaluation result to obtain a target ice thickness prediction model includes: When the prediction accuracy of the ice thickness prediction model is less than a preset prediction accuracy threshold, nonlinear transformation is performed on the independent variable of the ice thickness prediction model to obtain an adjusted ice thickness prediction model; Based on the target historical climate characteristic data and historical ice cover data, the adjusted ice cover thickness prediction model is refitted to obtain a target ice cover thickness prediction model.
8. A method for predicting ice thickness. It is characterized in that The method comprises: Obtain climate characteristic data of the target area within a preset time period; Taking the climate characteristic data as input, calling a preset ice thickness prediction model to obtain the ice thickness of the power facilities in the target area; Wherein, the preset ice thickness prediction model is obtained based on the ice thickness prediction model construction method described in any one of claims 1 to 7.
9. A device for constructing an ice thickness prediction model, It is characterized in that The device comprises: A historical data acquisition module, used to acquire historical ice coverage data and historical climate characteristic data of power facilities in the target area; A data screening module, for screening target historical climate characteristic data from the historical climate characteristic data based on the data collection time of the historical climate characteristic data; A model building module, used to build an initial ice thickness prediction model based on the target historical climate characteristic data and a multivariate linear regression model; The model fitting module is used to fit the initial ice thickness prediction model based on the target historical climate characteristic data and historical ice cover data, determine the regression coefficients of multiple independent variables in the initial ice thickness prediction model, and obtain the ice thickness prediction model.
10. An ice thickness prediction device, It is characterized in that The device comprises: A data acquisition module is used to acquire climate characteristic data of a target area within a preset time period; An ice thickness prediction module is used to use the climate characteristic data as input, call a preset ice thickness prediction model, and obtain the ice thickness of the power facilities in the target area; Wherein, the preset ice thickness prediction model is obtained based on the ice thickness prediction model construction method described in any one of claims 1 to 7.