Power transmission line icing type detection method and system based on hyperspectral satellite remote sensing

Abstract

The application discloses a power transmission line icing type detection method and system based on hyperspectral satellite remote sensing, which comprises the following steps: fusing geographical information and spectral remote sensing information of a power transmission line and obtaining characteristic standard values; calculating a joint index, obtaining key spectral characteristics related to an icing type according to the joint index, and screening an optimal feature combination; obtaining a training data set through sample data screening by the optimal feature combination, and training a power transmission line icing type prediction model; and predicting hyperspectral satellite remote sensing data based on the trained power transmission line icing type prediction model, and obtaining an icing type. The application realizes dynamic monitoring and long-time sequence analysis by using satellite remote sensing data, discovers icing conditions in time, realizes wide-area monitoring and rapid response to sudden weather events, and in addition, the hyperspectral satellite remote sensing does not need to install sensors on high-voltage power transmission lines, thereby avoiding safety risks of on-site construction.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power system operation and maintenance, and particularly relates to a power transmission line icing type detection method and system based on hyperspectral satellite remote sensing. BACKGROUND

[0002] As an important part of the power system, the stable operation of the power transmission line directly affects the continuity and reliability of power supply. However, the overhead power transmission line is exposed to a complex atmospheric environment for a long time, especially in winter or high-altitude areas, and often encounters icing phenomenon. Icing refers to the ice layer formed on the surface of the line due to low temperature, mainly including rain, fog and mixed rime. In recent years, with the global climate change, extreme weather events occur frequently, and the increase of extreme weather conditions such as sudden temperature drop, cold air invasion, ice rain and strong wind has aggravated the icing phenomenon of the power transmission line. Especially in mountainous areas, high altitudes and cold northern regions, the icing problem is particularly serious, and the icing of the power transmission line often causes power accidents and seriously affects the normal operation of the power system.

[0003] A power transmission line icing thickness monitoring method based on fiber grating sensors is proposed in the prior art. The fiber sensing technology is used to monitor the strain and temperature of the power transmission line in real time, and the data of the wind speed sensor is combined to calculate the line stress and infer the icing thickness. This method solves the problem of traditional icing detection relying on manual inspection, inaccuracy and difficulty in real-time monitoring, and is particularly suitable for monitoring needs under extreme weather conditions.

[0004] Although the above method has strong real-time performance and relatively high accuracy, it needs to lay multiple optical fibers and install multiple optical fiber sensors, which is high in construction and maintenance cost. Moreover, since the optical fiber sensor is installed on the power transmission line, maintenance and replacement need to be operated with power off, which has operation and maintenance difficulties. In addition, this method usually selects to arrange sensors at the connection between the power transmission line and the tower. This arrangement can monitor the change of the conductor stress between the towers, but cannot cover the entire power transmission line, and the monitoring range is limited. SUMMARY

[0005] In view of the above existing problems, the present application is proposed.

[0006] Therefore, the present application provides a power transmission line icing type detection method and system based on hyperspectral satellite remote sensing, which solves the problems of high construction and maintenance cost, difficult operation and maintenance, limited monitoring range and insufficient reliability under extreme weather conditions in the prior art.

[0007] To solve the above technical problems, the present application provides the following technical solutions:

[0008] In a first aspect, the present application provides a power transmission line icing type detection method based on hyperspectral satellite remote sensing, comprising: data fusion of geographical information and spectral remote sensing information of the power transmission line, and obtaining characteristic standard values; calculating a joint index based on the characteristic standard values, obtaining key spectral characteristics related to the icing type according to the joint index, and screening an optimal feature combination; obtaining a training data set by screening sample data through the optimal feature combination, and training a power transmission line icing type prediction model using the training data set; and predicting hyperspectral satellite remote sensing data based on the trained power transmission line icing type prediction model to obtain the icing type.

[0009] As a preferred scheme of the power transmission line icing type detection method based on hyperspectral satellite remote sensing, the acquisition of the geographical information and the spectral remote sensing information of the power transmission line comprises:

[0010] The input hyperspectral satellite remote sensing data is selected in near-infrared and short-wave infrared bands, and the input hyperspectral satellite remote sensing image is subjected to geometric correction, radiation calibration, and atmospheric correction preprocessing operations.

[0011] The hyperspectral satellite remote sensing data is labeled with an icing type according to the actual icing condition of the line, and sample cutting is performed.

[0012] The geographical information of the power transmission line includes altitude, slope, and slope direction information of the location of the power transmission line.

[0013] As a preferred scheme of the power transmission line icing type detection method based on hyperspectral satellite remote sensing, the obtaining of the characteristic standard values comprises:

[0014] The altitude, slope, and slope direction information of the location of the power transmission line calculated in combination with the remote sensing image elevation information is data fused with the spectral information.

[0015] The data fused features are standardized to obtain the characteristic standard values.

[0016] As a preferred scheme of the power transmission line icing type detection method based on hyperspectral satellite remote sensing, the calculation of the joint index comprises:

[0017] In the value range of the input feature X=(X1, X2,..., X n ), a small amplitude perturbation is performed on each variable, and the influence e i of each perturbation on the output on each track is calculated using a Latin hypercube sampling or random path design method.

[0018]

[0019] wherein, Δ represents the step size, Y represents the model output;

[0020] The calculation of the absolute mean μ* is:

[0021]

[0022] wherein, n represents the number of input features;

[0023] The calculation of the standard deviation σ is:

[0024]

[0025] The total variance V(Y) of the output Y is represented as:

[0026]

[0027] wherein, V i represents the variance contribution of a single input feature X i to Y, V ij represents the interaction effect of a single input feature X i to X j , V ijk represents the three-order interaction effect of a single input feature X i , X j , X k ;

[0028] The first-order sensitivity index S i is used to analyze the contribution of the input feature to the uncertainty of the output:

[0029]

[0030] The total effect sensitivity index S Ti represents the total contribution of the input feature X i , including its individual effect and interaction effect with other features, and is represented by the formula:

[0031]

[0032] As a preferred scheme of the ice type detection method for power transmission lines based on hyperspectral satellite remote sensing, the calculation of the joint index further includes:

[0033] A joint index I j is defined by combining the results of the Morris method and the Sobol sensitivity analysis, and is represented as:

[0034]

[0035] wherein, ω σ , ωs , respectively represent the weight of each index.

[0036] As a preferred scheme of the power transmission line icing type detection method based on hyperspectral satellite remote sensing provided by the application, the optimal feature combination screening comprises:

[0037] The sum of the mean and standard deviation of the joint index is selected as a threshold I, and I j The sample features of I are selected as the optimal feature combination.

[0038] As a preferred scheme of the power transmission line icing type detection method based on hyperspectral satellite remote sensing provided by the application, the optimal feature combination is used to screen sample data to obtain a training data set, and the training data set is used to train a power transmission line icing type prediction model based on a deep learning algorithm; the hyperspectral satellite remote sensing data is predicted based on the trained power transmission line icing type prediction model to obtain an icing type, wherein the icing type comprises three types of glaze, rime and mixed rime.

[0039] In a second aspect, the application provides a power transmission line icing type detection system based on hyperspectral satellite remote sensing, comprising:

[0040] A data fusion module is configured to fuse geographic information and spectral remote sensing information of the power transmission line to obtain feature standard values.

[0041] An optimal combination screening module is configured to calculate a joint index based on the feature standard values, obtain key spectral features related to the icing type according to the joint index, and screen an optimal feature combination.

[0042] A model training module is configured to screen sample data by using the optimal feature combination to obtain a training data set, and train a power transmission line icing type prediction model by using the training data set.

[0043] An icing type prediction module is configured to predict hyperspectral satellite remote sensing data based on the trained power transmission line icing type prediction model to obtain an icing type.

[0044] In a third aspect, the application provides an electronic device, comprising:

[0045] A memory and a processor.

[0046] The memory is configured to store computer executable instructions, and the processor is configured to execute the computer executable instructions, so as to implement the steps of the power transmission line icing type detection method based on hyperspectral satellite remote sensing.

[0047] In a fourth aspect, the present application provides a computer readable storage medium storing computer executable instructions, which, when executed by a processor, implement the steps of the power transmission line icing type detection method based on hyperspectral satellite remote sensing.

[0048] Compared with the prior art, the present application has the following beneficial effects: the present application provides a power transmission line icing type detection method and system based on hyperspectral satellite remote sensing, which is based on a deep learning algorithm, uses hyperspectral satellite remote sensing data and actual icing conditions of the power transmission line to train a model, obtains a power transmission line icing type prediction model, uses satellite remote sensing data to realize dynamic monitoring and long-time sequence analysis, discovers icing conditions in a timely manner, and performs wide-area monitoring and rapid response to sudden weather events. Hyperspectral data contains rich spectral information, and the data volume is sufficient, so the model detection precision is high. Hyperspectral satellite remote sensing is a non-contact monitoring method, which does not need to install sensors on the high-voltage power transmission line, thereby avoiding the safety risks of field construction. BRIEF DESCRIPTION OF DRAWINGS

[0049] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0050] Figure 1 The overall flow logic diagram of the power transmission line icing type detection method based on hyperspectral satellite remote sensing according to an embodiment of the present application is shown in the figure.

[0051] Figure 2 The overall flow block diagram of the power transmission line icing type detection method based on hyperspectral satellite remote sensing according to an embodiment of the present application is shown in the figure.

[0052] Figure 3 The basic convolution block Inception block diagram of the convolutional neural network GoogLeNet used in the power transmission line icing type detection method based on hyperspectral satellite remote sensing according to an embodiment of the present application is shown in the figure.

[0053] Figure 4 The GoogLeNet model diagram of the power transmission line icing type detection method based on hyperspectral satellite remote sensing according to an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0054] In order to make the above objectives, characteristics and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the protection scope of the present application.

[0055] Embodiment 1

[0056] Reference Figures 1-2 For an embodiment of the present application, a power transmission line icing type detection method based on hyperspectral satellite remote sensing is provided, as shown in Figure 1 Specifically includes the following steps:

[0057] S100: data fusion is performed on geographic information and spectral remote sensing information of the power transmission line, and a feature standard value is obtained;

[0058] S200: a joint index is calculated based on the feature standard value, key spectral features related to the icing type are obtained according to the joint index, and an optimal feature combination is screened;

[0059] S300: training data set is obtained by screening sample data through the optimal feature combination, the power transmission line icing type prediction model is trained by using the training data set, the hyperspectral satellite remote sensing data is predicted based on the trained power transmission line icing type prediction model, and the icing type is obtained.

[0060] It should be noted that the present application provides a power transmission line icing type detection method and system based on hyperspectral satellite remote sensing. Based on a deep learning algorithm, a power transmission line icing type prediction model is obtained by using hyperspectral satellite remote sensing data and actual icing conditions of the power transmission line for model training. Satellite remote sensing data is used to realize dynamic monitoring and long-time sequence analysis, to timely discover icing conditions, and to perform wide-area monitoring and rapid response to sudden weather events. Hyperspectral data contains rich spectral information, and the data amount is sufficient, so that the model detection precision is high. Hyperspectral satellite remote sensing is a non-contact monitoring method, which does not need to install sensors on the high-voltage power transmission line, thereby avoiding the safety risks of field construction.

[0061] In the embodiments of the present application, the above step S100 includes the following sub-steps A1-A3;

[0062] In A1: hyperspectral satellite remote sensing data is obtained and preprocessed;

[0063] In A2: elevation, slope and slope direction information of the position where the power transmission line is located is calculated in combination with remote sensing image elevation information;

[0064] In A3: The preprocessed hyperspectral satellite remote sensing data is fused with the calculated elevation, slope, and aspect information of the transmission line location, and characteristic standard values ​​are obtained.

[0065] Specifically, since the main absorption characteristics of ice are concentrated in the near-infrared (NIR) and short-wave infrared (SWIR) bands, the input hyperspectral satellite remote sensing data should be selected in the near-infrared and short-wave infrared bands, with a spectral resolution of not less than 10 nm and a spatial resolution of not less than 30 meters; Figure 2 The image shows that the input hyperspectral satellite remote sensing images are preprocessed using tools from a remote sensing image processing platform, including geometric correction, radiometric calibration, and atmospheric correction. Based on the actual icing conditions of the line, the hyperspectral satellite remote sensing data are labeled with icing type and the samples are cropped.

[0066] Specifically, the geographical information of transmission lines includes the elevation, slope, and aspect of the location of the transmission line;

[0067] Altitude information is the pixel value of the elevation information in a remote sensing image, which can be read directly.

[0068] Calculate the elevation change (gradient) of each raster cell in the X (east-west) and Y (north-south) directions using elevation information from remote sensing images:

[0069]

[0070] in, and It can be approximated by the finite difference method:

[0071]

[0072] Calculate the slope:

[0073]

[0074] Slope aspect indicates the direction of the surface slope (usually north is 0, clockwise is 2π). Calculating slope aspect requires the X and Y direction gradients of elevation changes.

[0075] Calculate the slope angle (in radians):

[0076] Aspect = arctan2(SlopeL) y -Slope x )

[0077] Specifically, the elevation, slope, and aspect information of the transmission line location, calculated from remote sensing image elevation information, will be fused with spectral information. The fused data features will then be standardized to obtain standard feature values. The formula for feature standardization is as follows:

[0078]

[0079] wherein, x i is a sample feature value, μ is a mean value of the sample feature, and σ is a standard deviation of the sample feature;

[0080] It should be noted that the above step S100 can not only integrate the advantages of multi-source data to provide more comprehensive and accurate power line environment information, but also lay a solid foundation for subsequent ice type prediction, ensure the reliability and representativeness of the feature standard value, and thus improve the accuracy and robustness of the entire prediction model.

[0081] In the embodiments of the present application, the above step S200 includes the following sub-steps B1-B2.

[0082] In B1: based on the feature standard value, a joint index is calculated using the Morris method and the Sobol sensitivity analysis index;

[0083] In B2: the key spectral features related to the ice type are obtained according to the joint index, and the optimal feature combination is selected;

[0084] Specifically, within the value range of the input feature X=(X1, X2,..., X n , a small perturbation is performed on each variable, the value of each variable is changed each time, and the other variables remain unchanged. The method of Latin hypercube sampling or random path design is used to calculate the influence of each perturbation on the output e i on each track:

[0085]

[0086] wherein, Δ represents the step size, and Y represents the model output;

[0087] The calculation of the absolute mean value μ * is as follows:

[0088]

[0089] wherein, n represents the number of input features;

[0090] The calculation of the standard deviation σ is as follows:

[0091]

[0092] The total variance Y(Y) of the output Y is represented as:

[0093]

[0094] wherein, V i represents a single input feature X iV's contribution to the variance of Y ij Representing a single input feature X i For X j The interaction effect, V ijk Representing a single input feature X i X j X k The third-order interaction effect;

[0095] Using the first-order sensitivity index S i The contribution of quantitative input features to output uncertainty is analyzed as follows:

[0096]

[0097] S i The value range is [0, 1], and the larger the value, the more significant X is. i The greater the impact on Y;

[0098] Total effect sensitivity index S Ti Indicates input feature X i The total contribution, including its individual effect and the interaction effect with other features, is expressed by the formula:

[0099]

[0100] A joint index I is defined by combining the results of the Morris method and the Sobol sensitivity analysis. j , is represented as:

[0101]

[0102] in, ω σ ω s , Each indicator represents its weight.

[0103] Specifically, selecting the optimal feature combination includes: choosing the sum of the mean and standard deviation of the joint indicators as a threshold I, and taking I... j The sample features > I are used as the optimal feature combination.

[0104] It should be noted that a more lenient screening criterion is to take the mean of the indicators and retain more features; a moderate screening criterion is to take the sum of the mean of the indicators and the standard deviation; a strict screening criterion is to take the sum of the mean of the indicators and twice the standard deviation and retain only the most important features; this embodiment adopts the moderate screening criterion.

[0105] It should be noted that the step S200 calculates the joint index based on the feature standard value, obtains the key spectral features related to the icing type according to the joint index, and screens out the optimal feature combination, which not only can effectively extract and highlight the features most influential to the icing type identification, but also can reduce data redundancy and noise, improve the representativeness of the features and the prediction accuracy of the model, thereby providing more reliable and efficient feature support for subsequent model training and icing type prediction.

[0106] In the embodiment of the present application, the step S300 screens the sample data by the optimal feature combination to obtain a training data set, trains the icing type prediction model of the power transmission line based on the deep learning algorithm and the training data set, and predicts the hyperspectral satellite remote sensing data based on the trained icing type prediction model of the power transmission line to obtain the icing type, wherein the icing type includes three types of glaze, rime and mixed rime.

[0107] It should be noted that the step S300 can not only ensure the high quality and representativeness of the training data, but also improve the training efficiency and prediction accuracy of the model, thereby providing more accurate and reliable model support for subsequent icing type prediction and effectively improving the accuracy and real-time performance of the icing monitoring.

[0108] The above is a schematic scheme of the method for detecting the icing type of the power transmission line based on the hyperspectral satellite remote sensing. It should be noted that the technical scheme of the system for detecting the icing type of the power transmission line based on the hyperspectral satellite remote sensing belongs to the same concept as the technical scheme of the method for detecting the icing type of the power transmission line based on the hyperspectral satellite remote sensing described above. The details of the technical scheme of the system for detecting the icing type of the power transmission line based on the hyperspectral satellite remote sensing in the embodiment are not described in detail, and can be referred to the description of the technical scheme of the method for detecting the icing type of the power transmission line based on the hyperspectral satellite remote sensing.

[0109] The system for detecting the icing type of the power transmission line based on the hyperspectral satellite remote sensing in the embodiment comprises:

[0110] The data fusion module is configured to fuse the geographic information and the spectral remote sensing information of the power transmission line to obtain the feature standard value.

[0111] The optimal combination screening module is configured to calculate the joint index based on the feature standard value, obtain the key spectral features related to the icing type according to the joint index, and screen the optimal feature combination.

[0112] The model training module is configured to screen the sample data by the optimal feature combination to obtain a training data set, and train the icing type prediction model of the power transmission line by using the training data set.

[0113] An icing type prediction module is configured to predict the hyperspectral satellite remote sensing data based on the trained power line icing type prediction model to obtain the icing type.

[0114] The above modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory in the computer device in software form, so as to be called and executed by the processor to perform the operations corresponding to the above modules.

[0115] The embodiment further provides an electronic device including a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. The processor of the computer device is configured 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 and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is configured to perform wired or wireless communication with an external terminal. The wireless communication can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The computer program is executed by the processor to implement the power line icing type detection method based on hyperspectral satellite remote sensing. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0116] The embodiment further provides a computer readable storage medium having a computer program stored thereon. The program is executed by the processor to implement the method proposed in the above embodiment.

[0117] The storage medium proposed in the embodiment and the method proposed in the above embodiment belong to the same inventive concept. The technical details not described in the embodiment can be referred to the above embodiment, and the embodiment has the same beneficial effects as the above embodiment.

[0118] Through the above description of the embodiments, those skilled in the art can clearly understand that the present application can be realized by means of software and necessary general hardware, and of course can also be realized by hardware, but in many cases the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a floppy disk, a read-only memory (ROM), a random access memory (RAM), a FLASH memory, a hard disk, or an optical disk, and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of various embodiments of the present application.

[0119] Embodiment 2

[0120] With reference to Figure 3 and Figure 4 , based on the previous embodiment, the present embodiment provides an application example of the power transmission line icing type detection method and system based on hyperspectral satellite remote sensing, which verifies and illustrates the technical effects used in the present method.

[0121] The present embodiment is applied to the inversion of power transmission line icing types in Guizhou Province in the past 25 years. First, the hyperspectral satellite remote sensing data of GF-5 in the past five years and the icing conditions of the power transmission line data in Guizhou Province are obtained. The spectral coverage of GF-5 is 400-2500 nm, the spectral resolution is 5-10 nm, and the spatial resolution is 30 meters. The related tools of the remote sensing image processing platform are used, and ENVI is used as an example to preprocess the hyperspectral satellite remote sensing data, including geometric correction, radiation calibration, and atmospheric correction. Then, according to the actual icing conditions of the line, the icing type label is marked on the hyperspectral remote sensing data, and finally the sample cutting is performed. According to the geographic elevation data (DEM), the altitude, slope, and slope direction information of the sample data are calculated and fused with the spectral information, and the fused features are standardized.

[0122] The results of the Morris method and Sobol sensitivity analysis are calculated to obtain the joint index I j :

[0123]

[0124] To ensure the calculation efficiency, the weights are set as: ω σ = 0.3, ω s = 0.2,

[0125] The distribution of the joint index is calculated, a moderate screening is selected, a threshold I is selected as the sum of the mean and standard deviation of the joint index, and I j The sample characteristics of I are selected as the optimal feature combination, the sample data is screened according to the optimal feature combination, and a training set is obtained; as shown in Figure 3 and Figure 4 The deep learning network used is GooLeNet. In GoogLeNet, the basic convolution block is called an Inception block.

[0126] Different scale features are extracted through parallel multi-scale convolution operations (1x1, 3x3 and 5x5 convolution) and maximum pooling operations; 1x1 convolution is used for dimension reduction to reduce the amount of calculation. The GooLeNet model has the advantages of small parameter quantity, high calculation efficiency and strong scalability: the Inception module can be adapted to different tasks by adjusting parameters. The GooLeNet model can solve the gradient disappearance problem of deep networks through auxiliary classifiers. The obtained training set is used to train the GooLeNet model, and a prediction model of the icing type of the power transmission line is obtained.

[0127] According to the power transmission line prediction model, the icing conditions of the power transmission line in Guizhou Province in the past 25 years are inversed, the icing types of the power transmission line in Guizhou Province in the past 25 years are obtained, and the inversion accuracy is 84.37%.

[0128] Therefore, the power transmission line icing type detection method based on hyperspectral satellite remote sensing provided in the embodiment can cover a large area range, especially suitable for the case where the power transmission lines are widely distributed, and overcomes the limitation of the limited monitoring range of ground sensors. Once imaging, it can cover long-distance power transmission lines, reducing the number of inspections. In addition, the satellite can pass through complex terrain areas, overcoming the limitations of mountainous, valley, forest and other terrains on the layout of ground sensors, especially suitable for monitoring power transmission lines in high-altitude or remote areas. The satellite has regular observation capability and can regularly obtain data according to the satellite orbit plan to realize dynamic monitoring and long-time sequence analysis, timely discover icing conditions, and is especially suitable for wide-area monitoring and rapid response to sudden weather events. Hyperspectral remote sensing is a non-contact monitoring method that does not require the installation of sensors on high-voltage power transmission lines, avoiding the safety risks of on-site construction, especially in severe weather conditions (such as strong winds, heavy snow and strong freezing weather). Hyperspectral remote sensing data contains rich spectral information (tens to hundreds of bands), which can capture subtle differences in different substances in the spectrum. The spectral reflection characteristics of the icing and non-icing conductors are different, and the hyperspectral data can identify such subtle changes to improve detection accuracy.

[0129] It should be noted that the above examples are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalent replaced without departing from the spirit and scope of the present application, and all of them should be covered in the scope of claims of the present application.

[0130] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt a computer program product in the form of being implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program codes. The solutions in the embodiments of the present application can be implemented in various computer languages.

[0131] The present application is described with reference to flowcharts and / or block diagrams according to the methods, devices (systems), and computer program products of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that carries out the functions specified in one or more blocks.

[0132] These computer program instructions can also be stored in a computer readable memory that can direct the computer or other programmable data processing devices to work in a specific manner, so that the instructions stored in the computer readable memory produce a manufactured product including instruction apparatus, which implements the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that carries out the functions specified in one or more blocks.

[0133] These computer program instructions can also be loaded into a computer or other programmable data processing device, so that a series of operation steps are performed on the computer or other programmable device to produce a computer implemented process, so that the instructions executed on the computer or other programmable device provide a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that carries out the functions specified in one or more blocks.

[0134] While the preferred embodiments of the application have been described, additional variations and modifications can be made to these embodiments by those skilled in the art once they have the benefit of the present disclosure without departing from the spirit and scope of the application. Accordingly, it is intended that such additions and modifications be included within the scope of the application. It is the following claims, including any amendments thereto, which define the scope of the application.

[0135] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described.

Claims

1. A method for detecting the icing type of a power transmission line based on hyperspectral satellite remote sensing, characterized in that, The method comprises the following steps: Data fusion is performed on geographical information and spectral remote sensing information of the power transmission line, and a characteristic standard value is obtained; A joint index is calculated based on the characteristic standard value, key spectral characteristics related to the icing type are obtained according to the joint index, and an optimal feature combination is screened; Training data set is obtained by screening sample data through the optimal feature combination, and a power transmission line icing type prediction model is trained by using the training data set; The icing type is obtained by predicting high-spectral satellite remote sensing data based on the trained power transmission line icing type prediction model. The calculation of the joint index comprises: In the input feature Small perturbations are applied to each variable within its value range, and the effect of each perturbation on the output is calculated using Latin hypercube sampling or random path design : wherein, denotes the step size, Y denotes the model output; absolute mean The calculation of the absolute mean is: Wherein, n represents the number of input features. Standard deviation is calculated as: Total variance of output is represented as: wherein, represents a single input feature a variance contribution to Y, represents a single input feature an interaction effect of , represents a single input feature , , a third order interaction effect of Using a first order sensitivity index quantifying the contribution of the input features to the output uncertainty is: Total effect sensitivity index represents the total contribution of the input features , including their individual effects and interaction effects with other features, is expressed as: 。 2. The power transmission line icing type detection method based on hyperspectral satellite remote sensing according to claim 1, characterized in that, The acquisition of the geographical information and the spectral remote sensing information of the power transmission line comprises: The near-infrared and short-wave infrared bands are selected from the input high-spectral satellite remote sensing data, and the input high-spectral satellite remote sensing image is preprocessed by geometric correction, radiation calibration and atmospheric correction; The high-spectral satellite remote sensing data is labeled with the icing type according to the actual icing condition of the line, and sample cutting is performed; The geographical information of the power transmission line comprises the altitude, slope and slope direction information of the location where the power transmission line is located.

3. The power transmission line icing type detection method based on hyperspectral satellite remote sensing according to claim 2, characterized in that, The acquisition of the characteristic standard value comprises: The altitude, slope and slope direction information of the location where the power transmission line is located calculated by combining the remote sensing image elevation information is data fused with the spectral information; The data fused features are standardized to obtain the characteristic standard value.

4. The power transmission line icing type detection method based on hyperspectral satellite remote sensing according to claim 3, characterized in that, The calculation of the joint index further comprises: A joint index is defined by combining the results of the Morris method and the Sobol sensitivity analysis, and is expressed as: wherein, respectively represent the weight of each index.

5. The power transmission line icing type detection method based on hyperspectral satellite remote sensing according to claim 4, characterized in that, The screening of the optimal feature combination comprises: The sum of the mean and the standard deviation of the combined index is selected as the threshold I, and the sample features of are taken as the optimal feature combination.

6. The power transmission line icing type detection method based on hyperspectral satellite remote sensing according to claim 5, characterized in that, Training data set is obtained by screening sample data through the optimal feature combination, and a power transmission line icing type prediction model is trained by using the training data set based on a deep learning algorithm; the icing type is obtained by predicting high-spectral satellite remote sensing data based on the trained power transmission line icing type prediction model, wherein the icing type comprises three types of glaze, rime and mixed rime.

7. A system for applying the method for detecting the icing type of a power transmission line based on hyperspectral satellite remote sensing according to any one of claims 1 to 6, characterized in that, The method comprises the following steps: A data fusion module is configured to perform data fusion on geographical information and spectral remote sensing information of the power transmission line, and obtain a characteristic standard value; An optimal combination screening module is configured to calculate a joint index based on the characteristic standard value, obtain key spectral characteristics related to the icing type according to the joint index, and screen an optimal feature combination; A model training module is configured to obtain training data set by screening sample data through the optimal feature combination, and train a power transmission line icing type prediction model by using the training data set; An icing type prediction module is configured to predict high-spectral satellite remote sensing data based on the trained power transmission line icing type prediction model, and obtain the icing type. 8.An electronic device, comprising: a memory and a processor; the memory is configured to store computer executable instructions, and the processor is configured to execute the computer executable instructions, and the computer executable instructions, when executed by the processor, implement the steps of the power transmission line icing type detection method based on high-spectral satellite remote sensing in any one of claims 1-6.

9. A computer readable storage medium storing computer executable instructions which, when executed by a processor, implement the steps of the method for detecting icing type of power transmission line based on hyperspectral satellite remote sensing according to any one of claims 1-6.

Images