Icing risk detection method and device, equipment and storage medium

Abstract

The embodiment of the invention provides an icing risk detection method and device, equipment and a storage medium. The method comprises the steps of obtaining elevation data and meteorological data of a target area, extracting basic topographic features of the target area and constructing icing influence indexes, and training a classification model for predicting icing risks based on the icing influence indexes and historical icing data. And inputting the elevation data and the meteorological data of the to-be-detected area into the classification model to obtain the icing risk level of the to-be-detected area. According to the icing influence indexes obtained through the basic topographic features and the meteorological data, the influence of the topographic features and the meteorological factors on icing can be incorporated into the classification model, the model is trained in combination with historical icing data, the prediction result of the to-be-detected area can be obtained, the prediction result can accurately reflect the icing risk level, and the prediction accuracy of the to-be-detected area is improved. Therefore, arrangement of icing monitoring points is facilitated, and the icing problem of the power transmission line is effectively solved.

Description

Technical Field

[0001] This application relates to the technical field of transmission lines, and particularly to an icing risk detection method, device, equipment and storage medium. Background Art

[0002] Under cold climate conditions, ice layers will form on the surface of transmission lines, and this phenomenon is usually called icing. Icing will have a serious impact on the safety and reliability of the power system. For the layout and maintenance of transmission lines, it is necessary to prevent and warn against icing risks.

[0003] In response to the icing problem faced by transmission lines, icing monitoring points need to be set up in the areas where the transmission lines are located. The difficulty in the site selection and layout of icing monitoring points lies in that different topographic regions have different degrees of influence on icing, and it is impossible to accurately evaluate the icing risks in different regions, resulting in difficulty in setting up monitoring points according to the icing risks. Summary of the Invention

[0004] Embodiments of this application provide an icing risk detection method, device, equipment and storage medium, so as to achieve the effect of improving the accuracy rate of icing risk assessment.

[0005] In a first aspect, embodiments of this application provide an icing risk detection method, including:

[0006] Obtain elevation data, meteorological data and historical icing data of a target area;

[0007] Extract basic topographic features according to the elevation data of the target area;

[0008] Determine an icing influence index according to the basic topographic features and the meteorological data of the target area;

[0009] Train a classification model based on the icing influence index and the historical icing data;

[0010] Input the elevation data and meteorological data of the area to be measured into the trained classification model to obtain the prediction result output by the classification model, and the prediction result represents the icing risk level of the area to be measured.

[0011] In a possible implementation manner, determining an icing influence index according to the basic topographic features and the meteorological data of the target area includes:

[0012] Calculate an icing topographic factor within the target area according to the basic topographic features, and the icing topographic factor includes the length of the windward slope and / or the valley depth ratio;

[0013] Calculate an icing meteorological factor according to the meteorological data of the target area;

[0014] Fuse the basic terrain features, the icing terrain factors, and the icing meteorological factors to obtain an icing impact index.

[0015] In a possible implementation manner, fusing the basic terrain features, the icing terrain factors, and the icing meteorological factors includes:

[0016] Conduct a correlation analysis on the basic terrain features, the icing terrain factors, and the icing meteorological factors, and set fusion weights according to the correlation analysis results;

[0017] Perform a weighted combination of the basic terrain features, the icing terrain factors, and the icing meteorological factors according to the fusion weights to obtain an icing impact index.

[0018] In a possible implementation manner, training a classification model based on the icing impact index and the historical icing data includes:

[0019] Generate a decision tree model according to the icing impact index and the historical icing data;

[0020] Use the elevation data and meteorological data of the target area as training samples and input them into the decision tree model, and adjust the model parameters of the decision tree model according to the output result of the decision tree model and the historical icing data. After the adjustment is completed, a classification model is obtained.

[0021] In a possible implementation manner, generating a decision tree model according to the icing impact index and the historical icing data includes:

[0022] Perform feature selection on the icing impact index according to the information gain ratio to determine the decision tree structure;

[0023] Optimize the decision tree structure through a pruning algorithm and construct a decision tree model corresponding to the optimized decision tree structure.

[0024] In a possible implementation manner, the historical icing data includes the icing thickness. Training the classification model based on the icing impact index and the historical icing data further includes:

[0025] Obtain the input value of the last layer of the classification model and calculate the first feature distribution according to the input value;

[0026] Calculate the second feature distribution according to the icing thickness;

[0027] Adjust the parameters of the classification model according to the difference between the first feature distribution and the second feature distribution.

[0028] In a possible implementation, before inputting the elevation data and meteorological data of the area to be measured into the trained classification model, it further includes:

[0029] Preprocess the elevation data and meteorological data of the area to be measured, and unify the preprocessed data into the same coordinate system.

[0030] In a second aspect, an icing risk detection device provided by an embodiment of the present application includes:

[0031] An acquisition module, configured to acquire elevation data, meteorological data, and historical icing data of a target area;

[0032] An extraction module, configured to extract basic terrain features according to the elevation data of the target area;

[0033] An index processing module, configured to determine an icing influence index according to the basic terrain features and the meteorological data of the target area;

[0034] A model training module, configured to train a classification model based on the icing influence index and the historical icing data;

[0035] A risk prediction module, configured to input the elevation data and meteorological data of the area to be measured into the trained classification model to obtain a prediction result output by the classification model, where the prediction result represents the icing risk level of the area to be measured.

[0036] In a third aspect, an electronic device provided by an embodiment of the present application includes: a memory, a processor;

[0037] The memory stores computer-executable instructions;

[0038] The processor executes the computer-executable instructions stored in the memory, so that the processor executes the above first aspect and / or various possible implementation manners of the first aspect.

[0039] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored, and when the computer-executable instructions are executed by a processor, they are used to implement the above first aspect and / or various possible implementation manners of the first aspect.

[0040] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the above first aspect and / or various possible implementation manners of the first aspect.

[0041] The icing risk detection method, device, equipment, and storage medium provided by the embodiments of the present application can extract the basic terrain features of the target area and construct icing impact indicators by obtaining the elevation data and meteorological data of the target area. Based on the icing impact indicators and historical icing data, a classification model for predicting icing risk can be trained, and then the elevation data and meteorological data of the area to be measured can be input into the classification model to obtain the icing risk level of the area to be measured. The icing impact indicators obtained from the basic terrain features and meteorological data can incorporate the impacts of terrain features and meteorological factors on icing into the classification model. By training the model in combination with historical icing data, a prediction result for the area to be measured can be obtained, which can accurately reflect the icing risk level, thereby helping to arrange icing monitoring points and effectively address the icing problem of transmission lines. Description of the Drawings

[0042] The drawings herein are incorporated into the specification and constitute a part of this specification, showing the embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0043] Figure 1 Schematic flow chart of an icing risk detection method provided by the present application Figure 1 ;

[0044] Figure 2 Schematic flow chart of an icing risk detection method provided by the present application Figure 2 ;

[0045] Figure 3 Schematic flow chart of a process for training a classification model provided by the present application Figure 1 ;

[0046] Figure 4 Schematic flow chart of a process for training a classification model provided by the present application Figure 2 ;

[0047] Figure 5 Schematic structural diagram of an icing risk detection device provided by the present application;

[0048] Figure 6 Schematic structural diagram of an electronic device provided by the present application.

[0049] Through the above drawings, the clear embodiments of the present application have been shown, and there will be more detailed descriptions hereinafter. These drawings and textual descriptions are not intended to limit the scope of the concept of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. Detailed Embodiments

[0050] Exemplary embodiments will be described in detail herein, and examples thereof are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0051] The icing problem of transmission lines refers to the phenomenon that ice layers form on the surfaces of conductors, tower structures, and other equipment of transmission lines under cold climate conditions. Icing increases the windward area of conductors and tower structures, thereby increasing the wind load and further increasing the stress on the structure. The ice layer also significantly increases the weight of conductors and tower structures, which may cause the conductors to sag, break, or the tower structures to collapse. In cold and humid regions, icing on transmission lines is the main cause of major accidents such as conductor breakage and tower collapse.

[0052] The generation and growth of icing are not only affected by climate but also restricted by topographical factors. Especially in some complex terrain environments such as hilly areas, the climate regulation effect of topography on local areas is often ignored, resulting in inaccurate assessment of icing risk, which in turn affects the layout of icing monitoring points.

[0053] Based on this, the present application proposes an icing risk detection method, device, equipment, and storage medium to solve the above technical problems.

[0054] The technical solution of the present application and how the technical solution of the present application solves the above technical problems will be described in detail below with specific embodiments. These several specific embodiments below can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below with reference to the accompanying drawings.

[0055] Figure 1 Flow schematic of an icing risk detection method provided by the present application Figure 1 as Figure 1 shown, the method includes:

[0056] Step S101, obtaining elevation data, meteorological data, and historical icing data of a target area.

[0057] Among them, the elevation data may refer to DEM (Digital Elevation Model) data, and the elevation value may represent the height information of a certain position in the area relative to a preset reference point (such as sea level).

[0058] In the embodiments of the present application, a target area can be selected from different terrain environments, and DEM data, meteorological data, historical icing data, etc. of this area can be obtained. For example, 10m resolution digital elevation model (DEM) data of the target area can be obtained through satellite remote sensing and lidar scanning, and the meteorological data of this area can be obtained from meteorological monitoring stations and other channels. The meteorological data may include information such as wind direction, temperature, and humidity of this area. The historical icing data may include data such as icing thickness for indicating the severity of icing in this area.

[0059] In some possible implementation manners, operations such as data cleaning and data complementation can be performed on the elevation data and meteorological data, and the elevation data and meteorological data can also be converted to the same coordinate system.

[0060] Step S102, extract basic terrain features according to the elevation data of the target area.

[0061] Among them, the basic terrain features may include slope, aspect, curvature, and elevation, etc. The slope can represent the degree of surface inclination, with the unit of degree. The aspect can represent the azimuth angle of the slope surface relative to the due north direction, with the unit of degree. The curvature may include profile curvature and plane curvature, which are used to characterize the concave and convex features of the terrain.

[0062] Specifically, tools such as GIS (Geographic Information System) software can be used to process the elevation data of the target area to extract basic terrain feature values such as slope, aspect, and curvature.

[0063] Exemplarily, for the slope, the neighborhood difference method can be used to calculate the elevation change rate of grid cells, and a continuously distributed value from 0° to 90° can be output; for the aspect, the eight-neighborhood analysis method can be used to determine the azimuth of the normal projection of the slope surface, and it can be classified as a shady slope (0° - 45°, 315° - 360°) and a sunny slope (135° - 225°); for the curvature, the profile curvature (reflecting runoff acceleration) and the plane curvature (characterizing the surface convergence feature) can be calculated through quadratic surface fitting.

[0064] Step S103, determine the icing influence index according to the basic terrain features and the meteorological data of the target area.

[0065] In the embodiments of the present application, first, it can be determined whether there are terrain factors in this area that affect icing according to the basic terrain features and meteorological data of the target area. The terrain factors may include valleys, windward slopes, etc. For example, valleys in this area can be identified according to data such as elevation values, or the main wind direction of this area can be determined first according to the meteorological data, and then the windward slope of this area can be identified according to data such as slope and aspect.

[0066] In some possible implementation manners, the basic terrain features, terrain factors, meteorological data, etc. can be combined in any form to construct a comprehensive icing influence index.

[0067] Step S104: Train a classification model based on the icing influence index and historical icing data.

[0068] In the embodiment of the present application, after obtaining the icing influence index, algorithms such as decision tree or XGBoost can be used based on the icing influence index to construct a classifier for determining the icing risk level. The classifier can output a prediction result representing the icing wind direction according to the input elevation data and meteorological data. Different icing risk levels can be divided based on the historical icing data as the true values to calibrate the parameters of the classifier. After calibration, a classification model capable of predicting the icing risk can be obtained based on the classifier.

[0069] Exemplarily, taking the XGBoost algorithm as an example, feature factors with a correlation coefficient greater than 0.8 with the icing thickness can be screened out from the basic terrain features, terrain factors, and meteorological data based on the mutual information method. The Box-Cox transformation can be performed on the continuous features to eliminate the skewed distribution. The objective function can be set as the weighted multi-class logarithmic loss, and a 3-fold weight can be applied to the high-risk level samples. Then, the Bayesian optimization is used to search for the best combination of hyperparameters (learning rate η ∈ [0.01, 0.3], tree depth d ∈ [5, 15]), and thus a classification model can be obtained.

[0070] Step S105: Input the elevation data and meteorological data of the area to be measured into the trained classification model to obtain the prediction result output by the classification model, and the prediction result represents the icing risk level of the area to be measured.

[0071] In the embodiment of the present application, the trained classification model can be deployed to the cloud GIS platform. The DEM data and meteorological data of the area to be measured are transmitted to the classification model, and the icing risk level of the area to be measured is determined by the classification model, so that the monitoring points can be set according to the icing risk level.

[0072] In some possible implementation manners, before inputting the elevation data and meteorological data of the area to be measured into the trained classification model, it further includes: preprocessing the elevation data and meteorological data of the area to be measured, and unifying the preprocessed data into the same coordinate system.

[0073] Among them, the preprocessing includes at least one of data denoising, data cleaning, and data completion.

[0074] Exemplarily, the Gaussian filtering algorithm can be adopted to denoise and smooth the DEM data of the area to be measured, the DEM data can be projected and transformed through a preset geographic coordinate system, and the meteorological data can be resampled to the same grid cells as the DEM through bilinear interpolation to achieve the spatial registration of the DEM and the meteorological data; to align the time series, the sliding window mean filtering (window length 24 hours) can also be implemented on the meteorological data to synchronize its time resolution with the icing monitoring data to the daily scale; in addition, based on the 3σ criterion (the Pauta criterion), the abnormal DEM elevation points are removed, and the Kriging spatial interpolation method is used to complement the missing data.

[0075] In the above embodiment, the elevation data and meteorological data of the target area are obtained, the basic terrain features of the target area can be extracted and the icing influence index can be constructed. Based on the icing influence index and historical icing data, a classification model for predicting the icing risk can be trained, and then the elevation data and meteorological data of the area to be measured are input into the classification model to obtain the icing risk level of the area to be measured. Through the icing influence index obtained from the basic terrain features and meteorological data, the influence of terrain features and meteorological factors on icing can be incorporated into the classification model, and the model is trained in combination with historical icing data to obtain the prediction result for the area to be measured. This prediction result can accurately reflect the icing risk levels of different terrain areas, thus helping to arrange icing monitoring points and effectively deal with the icing problem of transmission lines.

[0076] Figure 2 The flow chart of an icing risk detection method provided by this application Figure 2 , such as Figure 2 shown, determining the icing influence index according to the basic terrain features and the meteorological data of the target area may include:

[0077] Step S201, calculating the icing terrain factor within the target area according to the basic terrain features.

[0078] Among them, the icing terrain factor includes the length of the windward slope and / or the valley depth ratio.

[0079] The length of the windward slope can represent the force path length of the slope wind. Several sample points can be selected and the length of the windward slope can be calculated through the following formula (1).

[0080] (1)

[0081] Among them, L is the length of the windward slope, is the slope height, is the slope, and n is the number of sample points.

[0082] The valley depth ratio can represent the ratio of the depth to the width of the valley and can be calculated through the following formula (2).

[0083] (2)

[0084] Wherein, R is the valley depth ratio, and are respectively the maximum elevation value and the minimum elevation value of the valley, and D is the valley width.

[0085] Step S202, calculate the icing meteorological factors according to the meteorological data of the target area.

[0086] Wherein, the icing meteorological factors may include the temperature factor F T and the humidity factor F RH . The temperature factor F T can be obtained through the following formula (3), and the humidity factor F RH can be obtained through the following formula (4).

[0087] (3)

[0088] Wherein, T is the temperature, σ is the width parameter of the temperature distribution, and σ 1 =2, σ 2 =6.

[0089] (4)

[0090] Wherein, RH can represent the relative humidity, and k is the exponential parameter that controls the attenuation rate when RH is lower than 90, and k = 3 can be taken.

[0091] Step S203, fuse the basic terrain features, icing terrain factors and icing meteorological factors to obtain the icing influence index.

[0092] In some possible implementation manners, perform a correlation analysis on the basic terrain features, icing terrain factors and icing meteorological factors, and set the fusion weights according to the correlation analysis results; perform a weighted combination of the basic terrain features, icing terrain factors and icing meteorological factors according to the fusion weights to obtain the icing influence index.

[0093] Exemplarily, a matrix including basic terrain features (such as slope, aspect and curvature, etc.), icing terrain factors (such as the length of the windward slope and the valley depth ratio) and icing meteorological factors can be established through the principal component analysis method, calculate the covariance matrix and determine the eigenvectors; select the principal components with the cumulative contribution rate greater than the preset threshold to construct the icing influence index.

[0094] The inventor has discovered through research that the severity of icing risk is not only proportional to the length of the windward slope but also to the value of the valley depth ratio. Topographic factors such as the length of the windward slope and the valley depth ratio that have a greater impact on icing can be extracted from the basic topographic features such as the slope and aspect of the target area, as well as meteorological data. Based on this, an icing impact index can be constructed to more accurately determine the icing risk level.

[0095] In one embodiment, as Figure 3 shown, training a classification model based on the icing impact index and historical icing data includes:

[0096] Step S301, generating a decision tree model according to the icing impact index and historical icing data.

[0097] In some possible implementation manners, feature selection can be performed on the icing impact index according to the information gain ratio to determine the decision tree structure; the decision tree structure is optimized through a pruning algorithm, and a decision tree model corresponding to the optimized decision tree structure is constructed.

[0098] Exemplarily, splitting attribute selection can be performed based on the information gain ratio. For each candidate feature, its information gain ratio is calculated, and the feature with the largest information gain ratio is selected as the splitting basis for the current node to eliminate the bias caused by the relatively large number of feature values, and the decision tree structure is initially determined; the decision tree structure is optimized using the Cost-Complexity Pruning algorithm, and the optimal pruning strength is determined through cross-validation, and the subtree with an incremental misclassification cost less than the complexity penalty is removed.

[0099] Among them, the optimization process includes calculating the complexity cost of the subtree, and the calculation process can refer to formula (5).

[0100] R α (T)=R(T)+α×|T| (5)

[0101] In the above formula (5), R(T) is the misclassification cost, |T| is the number of leaf nodes, and α is the penalty coefficient.

[0102] Step S302, using the elevation data and meteorological data of the target area as training samples and inputting them into the decision tree model, and adjusting the model parameters of the decision tree model according to the output result of the decision tree model and historical icing data. After the adjustment is completed, a classification model is obtained.

[0103] In the embodiments of the present application, feature data such as elevation data and meteorological data can be normalized, and the processed data is used as sample data and divided into a training set, a validation set, and a test set, and the performance of the classification model is evaluated using the accuracy rate and the F1 score (an index to measure the accuracy of the model).

[0104] The classification model can output different icing risk levels, such as low risk, medium risk, high risk, and extremely high risk, etc. Optionally, the classification model can distinguish different risk levels by calculating the probability of icing and / or the icing thickness in the intermediate layer calculation area.

[0105] Through the above embodiments, by clarifying the feature selection mechanism, the defect that the traditional information gain favors multi-valued features can be overcome, and by quantifying the cost complexity calculation in the pruning algorithm, the parameter optimization process of the classification model can be strengthened, and the accuracy of icing risk prediction can be improved.

[0106] In one embodiment, the historical icing data includes the icing thickness, such as Figure 4 As shown, training the classification model based on the icing impact index and historical icing data further includes:

[0107] Step S401, obtain the input value of the last layer of the classification model, and calculate the first feature distribution according to the input value.

[0108] The classification model can include multiple intermediate layers and a final output layer. Among them, the intermediate layer can be used to process the features extracted by the model, and the intermediate layer finally inputs the processed features into the last layer (i.e., the output layer) of the model, and the output layer can be used to output the icing risk level.

[0109] Step S402, calculate the second feature distribution according to the icing thickness.

[0110] Step S403, adjust the parameters of the classification model according to the difference between the first feature distribution and the second feature distribution.

[0111] In this embodiment, after inputting the elevation data and meteorological data of the target area into the classification model, the classification model can predict the icing thickness of the target area in the intermediate layer, and the last layer of the classification model can determine which icing risk level the target area belongs to according to the icing thickness predicted by the intermediate layer.

[0112] Exemplarily, the first feature distribution of the icing thickness predicted by the intermediate layer can be determined according to the input value of the last layer of the classification model, and the second feature distribution can be calculated in combination with the actual icing thickness in the historical icing data, so as to optimize the parameters of the classification model according to the distribution difference between the two.

[0113] In some possible implementation manners, the KL (Kullback-Leibler Divergence, also known as relative entropy or information divergence) divergence loss function can be introduced to measure the consistency between the first feature distribution and the Gaussian distribution law.

[0114] Exemplarily, the difference between the first feature distribution and the second feature distribution can be calculated by the KL divergence loss function. When the value of the KL divergence loss function is large, it indicates that the degree of difference between the two is large; when the value of the KL divergence is small, it indicates that the difference between the two is small. By backpropagating the calculated KL divergence value, the model weights can be optimized, making the feature distribution of the model closer to the Gaussian distribution, thereby improving the prediction accuracy of the model.

[0115] Figure 5 The following is a schematic structural diagram of an icing risk detection device provided by this application, as Figure 5 shown, the icing risk detection device 500 provided in this embodiment includes:

[0116] An acquisition module 501, configured to acquire elevation data, meteorological data, and historical icing data of a target area;

[0117] An extraction module 502, configured to extract basic terrain features according to the elevation data of the target area;

[0118] An index processing module 503, configured to determine icing influence indexes according to the basic terrain features and the meteorological data of the target area;

[0119] A model training module 504, configured to train a classification model based on the icing influence indexes and the historical icing data;

[0120] A risk prediction module 505, configured to input the elevation data and meteorological data of a to-be-detected area into the trained classification model to obtain a prediction result output by the classification model, where the prediction result characterizes the icing risk level of the to-be-detected area.

[0121] In a possible implementation manner, the index processing module 503 is further configured to: calculate an icing terrain factor in the target area according to the basic terrain features, where the icing terrain factor includes a windward slope length and / or a valley depth ratio; calculate an icing meteorological factor according to the meteorological data of the target area; and fuse the basic terrain features, the icing terrain factor, and the icing meteorological factor to obtain an icing influence index.

[0122] In a possible implementation manner, the index processing module 503 is further configured to: perform a correlation analysis on the basic terrain features, the icing terrain factor, and the icing meteorological factor, and set a fusion weight according to the correlation analysis result; and perform a weighted combination of the basic terrain features, the icing terrain factor, and the icing meteorological factor according to the fusion weight to obtain an icing influence index.

[0123] In a possible implementation, the model training module 504 is further configured to: generate a decision tree model according to the icing impact index and the historical icing data; use the elevation data and meteorological data of the target area as training samples to input into the decision tree model, and adjust the model parameters of the decision tree model according to the output result of the decision tree model and the historical icing data. After the adjustment is completed, a classification model is obtained.

[0124] In a possible implementation, the model training module 504 is further configured to: perform feature selection on the icing impact index according to the information gain ratio to determine the decision tree structure; optimize the decision tree structure through a pruning algorithm, and construct a decision tree model corresponding to the optimized decision tree structure.

[0125] In a possible implementation, the model training module 504 is further configured to: obtain the input value of the last layer of the classification model, and calculate the first feature distribution according to the input value; calculate the second feature distribution according to the icing thickness; adjust the parameters of the classification model according to the difference between the first feature distribution and the second feature distribution.

[0126] In a possible implementation, the risk prediction module 505 is further configured to: preprocess the elevation data and meteorological data of the area to be measured, and unify the preprocessed data into the same coordinate system.

[0127] The icing risk detection device provided in this embodiment can execute the method provided in the above method embodiment, and its implementation principle and technical effect are similar, which will not be elaborated here in this embodiment.

[0128] Figure 6 The following is a schematic structural diagram of an electronic device provided by the present application. As Figure 6 shown, the electronic device 60 provided in this embodiment includes: at least one processor 601 and a memory 602. Optionally, the device 60 further includes a communication component 603. Among them, the processor 601, the memory 602, and the communication component 603 are connected through a bus 604.

[0129] In a specific implementation process, at least one processor 601 executes the computer execution instructions stored in the memory 602, so that at least one processor 601 executes the above method.

[0130] The specific implementation process of the processor 601 can refer to the above method embodiment, and its implementation principle and technical effect are similar, which will not be elaborated here in this embodiment.

[0131] In the above embodiments, it should be understood that the processor may be a central processing unit (CPU for short), or may also be other general-purpose processors, digital signal processors (DSP for short), application specific integrated circuits (ASIC for short), etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The steps of the method disclosed in combination with the invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules in the processor.

[0132] The memory may include a high-speed memory (Random Access Memory, RAM), and may also include non-volatile memory (Non-volatile Memory, NVM), such as at least one disk memory.

[0133] The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience in representation, the buses in the drawings of this application are not limited to only one bus or one type of bus.

[0134] This application also provides a computer program product, including a computer program, which implements the above method when executed by a processor.

[0135] This application also provides a computer-readable storage medium, in which computer-executable instructions are stored, and when the processor executes the computer-executable instructions, the above method is implemented.

[0136] The above-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, a magnetic disk or an optical disc. The readable storage medium can be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0137] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be part of the processor. The processor and the readable storage medium can be located in an Application Specific Integrated Circuit (ASIC). Of course, the processor and the readable storage medium can also exist as discrete components in a device.

[0138] The division of units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the couplings or direct couplings or communication connections shown or discussed among each other can be indirect couplings or communication connections through some interfaces, devices or units, and can be in electrical, mechanical or other forms.

[0139] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0140] In addition, in each embodiment of the present invention, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0141] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it 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. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of each embodiment of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks or optical discs that can store program codes.

[0142] Those of ordinary skill in the art will understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions. The foregoing program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments; and the foregoing storage medium includes: various media such as ROM, RAM, magnetic disks, or optical disks that can store program codes.

[0143] Finally, it should be noted that those skilled in the art will readily conceive of other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. The present invention is intended to cover any variations, uses, or adaptations of the present invention, which follow the general principles of the present invention and include known common knowledge or conventional technical means in the technical field not disclosed by the present invention. It is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present invention is only limited by the appended claims.

Claims

1. A method for detecting ice risk, characterized in that: include: Obtain elevation data, meteorological data and historical ice cover data of the target area; extracting basic terrain features based on elevation data of the target area; Determining an ice coverage impact index based on the basic terrain features and the meteorological data of the target area; Training a classification model based on the icing impact index and the historical icing data; The elevation data and meteorological data of the area to be measured are input into the trained classification model to obtain the prediction result output by the classification model, and the prediction result represents the ice risk level of the area to be measured.

2. The method according to claim 1, characterized in that The step of determining the ice coverage impact index according to the basic terrain features and the meteorological data of the target area includes: Calculating an ice-covered terrain factor in the target area according to the basic terrain features, the ice-covered terrain factor including a windward slope length and / or a valley depth ratio; Calculating an icing meteorological factor according to the meteorological data of the target area; The basic terrain features, the ice-covered terrain factors and the ice-covered meteorological factors are integrated to obtain an ice-covered impact index.

3. The method according to claim 2, characterized in that The fusing of the basic terrain features, the ice-covered terrain factors and the ice-covered meteorological factors comprises: Performing correlation analysis on the basic terrain features, the ice-covered terrain factors and the ice-covered meteorological factors, and setting fusion weights according to the correlation analysis results; The basic terrain features, the ice-covered terrain factors and the ice-covered meteorological factors are weighted and combined according to the fusion weights to obtain an ice-covered impact index.

4. The method according to any one of claims 1 to 3, characterized in that The training of the classification model based on the icing impact index and the historical icing data includes: Generate a decision tree model according to the icing impact index and the historical icing data; The elevation data and meteorological data of the target area are input into the decision tree model as training samples, and the model parameters of the decision tree model are adjusted according to the output results of the decision tree model and the historical ice cover data. After the adjustment is completed, a classification model is obtained.

5. The method according to claim 4, characterized in that The generating a decision tree model according to the icing impact index and the historical icing data includes: Performing feature selection on the ice coverage impact index according to the information gain rate to determine a decision tree structure; The decision tree structure is optimized by a pruning algorithm, and a decision tree model corresponding to the optimized decision tree structure is constructed.

6. The method according to claim 4, characterized in that The historical ice coverage data includes ice thickness, and the training of the classification model based on the ice coverage impact index and the historical ice coverage data further includes: Obtaining an input value of the last layer of the classification model, and calculating a first feature distribution according to the input value; Calculating a second characteristic distribution according to the ice thickness; According to the difference between the first feature distribution and the second feature distribution, the parameters of the classification model are adjusted.

7. The method according to any one of claims 1 to 3, characterized in that Before the elevation data and meteorological data of the area to be tested are input into the trained classification model, it also includes: The elevation data and meteorological data of the area to be measured are preprocessed, and the preprocessed data are unified into the same coordinate system.

8. An icing risk detection device, characterized in that: include: An acquisition module is used to obtain elevation data, meteorological data and historical ice cover data of the target area; An extraction module, used for extracting basic terrain features according to the elevation data of the target area; An index processing module, used for determining an ice coverage impact index according to the basic terrain features and the meteorological data of the target area; A model training module, used for training a classification model based on the icing impact index and the historical icing data; The risk prediction module is used to input the elevation data and meteorological data of the area to be tested into the trained classification model to obtain the prediction result output by the classification model, and the prediction result represents the ice risk level of the area to be tested.

9. An electronic device, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.

Images