Power grid engineering scene vegetation growth model method and system

By identifying power grid facilities and soil areas in power grid engineering scenarios, and using prediction models and image processing technology to quickly predict vegetation growth distribution, the problem of slow processing speed caused by large amount of remote sensing image data is solved, the efficiency of power grid facilities management is improved, and the power grid is guaranteed Operational safety and stability.

CN118941954APending Publication Date: 2024-11-12STATE GRID SHANGHAI ELECTRIC POWER DESIGN
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410983510.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

In the existing technology, the large amount of remote sensing image data is caused by the need to perform a lot of image processing and analysis when predicting the growth of vegetation in a certain period of the future, resulting in slow processing speed, reducing the efficiency of power grid facilities management, and affecting the safety and stability of power grid operation. .

Method used

By obtaining remote sensing images of power grid engineering scenarios, identifying the location and soil area of ​​the power grid facility, demarcate a designated area centered on the power grid facility, predicting future vegetation growth distribution using a trained prediction model, and assessing its impact on the power grid facility. Combined with YOLO object detection model and support vector machine for image classification and prediction, NDVI values ​​are used to calculate the degree of vegetation coverage, and quickly and accurately predict vegetation growth distribution.

Benefits of technology

It improves the management efficiency of power grid facilities, reduces the impact of vegetation growth on power grid facilities, ensures the safety and stability of power grid operation, and significantly reduces the computational complexity and time cost.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118941954A_ABST
    Figure CN118941954A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of power grid engineering scene vegetation growth prediction, and particularly discloses a power grid engineering scene vegetation growth model method and system, and the method mainly comprises the steps: obtaining a remote sensing image of a power grid engineering scene; identifying a power grid facility position and a soil area in the real-time remote sensing image; delimiting a specified range area which takes the position of the power grid facility as the center and is a soil area on the real-time remote sensing image; inputting the remote sensing image into the trained prediction model to obtain predicted vegetation growth distribution in the specified time and the specified range in the future; and evaluating the influence of the predicted vegetation growth distribution on the power grid facilities. According to the method, the vegetation growth distribution in the specified time and the specified range in the future can be quickly and accurately predicted, the efficiency of power grid facility management is improved, the influence of vegetation growth on power grid facilities is reduced, and the safety and the stability of power grid operation are guaranteed.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of vegetation growth prediction in power grid engineering scenarios, and in particular to a vegetation growth model method for power grid engineering scenarios and a vegetation growth model system for power grid engineering scenarios. Background Art

[0002] The growth of vegetation may entangle electric poles and lines, causing short circuits, disconnections, and even fires. In addition, fallen leaves and branches of trees may accumulate on electrical equipment, increasing the risk of equipment failure. Vegetation will also affect the heat dissipation of power grid facilities, causing equipment overheating and reducing operating efficiency. At the same time, the growth of vegetation will also affect the observation and maintenance of power grid facilities, increasing maintenance costs. Therefore, it is necessary to reasonably plan the growth of vegetation in power grid projects, take measures to control the impact of vegetation on power grid facilities, and ensure the safe and stable operation of the power grid.

[0003] Remote sensing images, GIS technology, etc. are used to predict the growth of vegetation in a certain period of time in the future. Based on the prediction results, combined with information such as the type, location, and structure of the power grid facilities, it can be determined whether the vegetation will have an impact on the power grid facilities. However, the amount of remote sensing image data is huge, and a large amount of image processing and analysis is required, which makes the processing speed very slow, reducing the efficiency of power grid facility management and may affect the normal operation of the power grid. Summary of the invention

[0004] The present application provides a vegetation growth model method and system for a power grid engineering scenario, thereby solving the problem in the prior art that the amount of remote sensing image data is huge, and a large amount of image processing and analysis is required when predicting the growth of vegetation in a certain period in the future, resulting in a very slow processing speed. The application improves the efficiency of power grid facility management, reduces the impact of vegetation growth on power grid facilities, and ensures the safety and stability of power grid operation.

[0005] In the first aspect, the present application provides a vegetation growth model method for a power grid engineering scenario, which includes the following steps: obtaining remote sensing images of the power grid engineering scenario, including real-time remote sensing images and historical remote sensing images; identifying the location of power grid facilities and soil areas in the real-time remote sensing images; delineating a designated range area centered on the location of the power grid facilities and the soil area on the real-time remote sensing images; inputting the remote sensing images into a trained prediction model to obtain the predicted vegetation growth distribution within the specified future time and specified range area; and evaluating the impact of the predicted vegetation growth distribution on the power grid facilities.

[0006] By demarcating a designated range area centered on the location of the power grid facility and the soil area, the vegetation area that may be affected around the power grid facility can be efficiently defined, and the vegetation growth distribution within the designated time and designated range in the future can be quickly and accurately predicted, thereby improving the efficiency of power grid facility management, reducing the impact of vegetation growth on power grid facilities, and ensuring the safety and stability of power grid operation.

[0007] Furthermore, a YOLO target detection model is trained using the remote sensing image with marked power grid facilities, and the real-time remote sensing image is input into the YOLO target detection model to obtain a remote sensing image with marked power grid facility locations and bounding boxes.

[0008] Furthermore, sample points representing soil areas and non-soil areas are selected in the remote sensing image, the collected sample points are used to train a support vector machine, and the trained classifier is used to perform classification prediction on the remote sensing image to obtain the classification results of soil areas and non-soil areas.

[0009] Furthermore, delineating a designated range area centered on the location of the power grid facility and being a soil area specifically includes: selecting a circular area centered on the power grid facility and having a radius of R0+k*I(s); selecting a soil area within the circular area to obtain the designated range area; wherein R0 represents a basic radius, and k represents an adjustment coefficient for adjusting the radius size according to the degree to which the power grid facility is susceptible to vegetation. Where α and β are parameters that control the shape of the function, s represents the type of power grid facilities, I(s)∈(0,1), I(s) is close to 1, indicating that the power grid facilities are easily affected by vegetation, and I(s) is close to 0, indicating that the power grid facilities are not easily affected by vegetation.

[0010] Furthermore, the remote sensing image has an infrared band and a visible light band, and NDVI calculation is performed on the infrared band and the red band in the remote sensing image to obtain the NDVI value of each pixel; the remote sensing image with the NDVI value is input into a trained prediction model, and the remote sensing image with the NDVI value predicted in the future within a specified time and a specified range is output, and the predicted vegetation growth distribution in the future within the specified time and a specified range is obtained based on the real-time remote sensing image and the predicted remote sensing image with the NDVI value.

[0011] Furthermore, based on the real-time remote sensing image and the predicted remote sensing image with NDVI value, the predicted vegetation growth distribution in the specified future time and specified range area is obtained, which specifically includes: defining the vegetation value VR of each pixel point in the real-time remote sensing image ij , VR ij = {vr ij}, where v ijRepresents the vegetation of the jth growth level of the i-th type in the real-time remote sensing image; based on the NDVI value, the vegetation value VP of each pixel in the remote sensing image with NDVI value within the specified future time and specified range is calculated and predicted. ij , VP ij for: In the formula, NDVI rk Represents the NDVI value of the kth pixel in the real-time remote sensing image, NDVI pk Represents the predicted NDVI value of the kth pixel in the remote sensing image, vp ij Represents the predicted vegetation of the jth growth level of the ith type in the remote sensing image.

[0012] Further, vegetation types include: trees, shrubs and grass.

[0013] Furthermore, evaluating the impact of the predicted vegetation growth distribution on the power grid facilities specifically includes: traversing each power grid facility, determining whether the power grid facility is covered by vegetation, and when the power grid facility is covered by vegetation, determining the degree of impact of the power grid facility. c For: Impac c =(w t V t +w s V s +

[0014] w g V g )×C×F1; where V t 、V s and V g Represents the growth degree of trees, shrubs and grass respectively, w t 、w s and w g Respectively represent the weight coefficients of the impact of trees, shrubs and grasslands on power grid facilities, C represents the coverage of vegetation on power grid facilities, and F1 represents the comprehensive factor; when the power grid facilities are not covered by vegetation, the impact degree of the power grid facilities is d For: Impac d =(w t V t +w s V s +w g V g )×D×F2; where D represents the closest distance to the power grid facilities and F2 represents the comprehensive factor.

[0015] Furthermore, the coverage degree C of vegetation on power grid facilities is: Where N P Represents the number of pixels occupied by power grid facilities, NV Represents the number of pixels occupied by vegetation among the pixels occupied by power grid facilities.

[0016] In the second aspect, the present application provides a vegetation growth model system for a power grid engineering scenario, which adopts the vegetation growth model method for a power grid engineering scenario as described in the first aspect, and includes: an image acquisition module, a feature recognition module, a range selection module, a vegetation prediction module and an impact assessment module.

[0017] The image acquisition module is used to acquire remote sensing images of power grid engineering scenes, including real-time remote sensing images and historical remote sensing images; the feature recognition module is used to identify the location of power grid facilities and soil areas in the real-time remote sensing images; the range selection module is used to delineate a designated range area centered on the location of the power grid facilities and the soil area on the real-time remote sensing images; the vegetation prediction module is used to input the remote sensing images into a trained prediction model to obtain the predicted vegetation growth distribution within the specified future time and specified range area; the impact assessment module is used to assess the impact of the predicted vegetation growth distribution on the power grid facilities.

[0018] In a third aspect, the present application also provides a vegetation growth model device for a power grid engineering scenario, which includes a memory and a processor; the memory is used to store a computer program; the processor is used to implement the steps of the vegetation growth model method for a power grid engineering scenario as in the first aspect when executing the computer program.

[0019] The technical solution provided by this application has at least the following technical effects or advantages:

[0020] By demarcating a designated range area centered on the location of the power grid facility and which is the soil area, the predicted vegetation growth distribution at a specified time in the future is evaluated, and the impact of the predicted vegetation growth distribution on the power grid facility is evaluated. This effectively solves the problem in the prior art that the amount of remote sensing image data is huge, and a large amount of image processing and analysis is required when predicting the growth of vegetation in a certain period in the future, resulting in a very slow processing speed. The method can quickly and accurately predict the vegetation growth distribution in the specified time and range in the future, thereby improving the efficiency of power grid facility management, reducing the impact of vegetation growth on power grid facilities, and ensuring the safety and stability of power grid operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a flow chart of the vegetation growth model method for the power grid engineering scenario in this application;

[0022] Figure 2 A flow chart for evaluating the impact of predicted vegetation growth distribution on the power grid facilities in this application. DETAILED DESCRIPTION

[0023] In order to solve the problem that a large amount of image processing and analysis is required when predicting vegetation growth, which causes the processing speed to become very slow, the vegetation growth model method proposed in this application can quickly and accurately predict the vegetation growth distribution within a specified time and specified range in the future, thereby improving the efficiency of power grid facility management, reducing the impact of vegetation growth on power grid facilities, and ensuring the safety and stability of power grid operation.

[0024] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0025] Embodiment 1

[0026] like Figure 1-Figure 2 As shown, this embodiment provides a vegetation growth model method for a power grid engineering scenario, which includes the following steps:

[0027] S100. Acquire remote sensing images of power grid engineering scenes, including real-time remote sensing images and historical remote sensing images.

[0028] It is an image of the earth's surface taken by platforms such as satellites, airplanes or drones. Using remote sensing images, we can intuitively see the location distribution of power grid facilities and the vegetation conditions in the surrounding environment. Historical remote sensing images record the vegetation conditions in a certain period of time in the past. By comparing real-time and historical images, we can understand the trend of vegetation changes over time, so as to better predict the impact of future vegetation growth on power grid facilities. If historical images show that the vegetation coverage rate in the area is increasing year by year, and real-time images show that the vegetation is close to the power grid facilities, then we can predict that vegetation may hinder or block the power grid facilities in the future, and corresponding measures need to be taken for maintenance and management.

[0029] S200. Identify power grid facility locations and soil areas in real-time remote sensing images.

[0030] Although not all soil areas are suitable for vegetation growth, soil is an indispensable foundation for vegetation growth. Therefore, the soil area represents the environment in which vegetation can potentially grow. By comprehensively considering the location of power grid facilities and the soil area, we can more accurately predict the impact of future vegetation on power grid facilities and take appropriate measures for planning and management.

[0031] S210. Use the remote sensing image with marked power grid facilities to train a YOLO target detection model, input the real-time remote sensing image into the YOLO target detection model, and obtain the remote sensing image with marked power grid facility location and bounding box.

[0032] When training the YOLO target detection model, the existing YOLOv3 or YOLOv4 model structure is used. According to the categories and the number of bounding boxes in the data set, the input layer and output layer of the model are adjusted. The power grid facilities in the image are annotated using annotation tools (such as LabelImg, Labelbox, etc.), and the bounding box and category information of each target are generated. The constructed YOLO model is trained using the annotated remote sensing image data set. The detection accuracy of the model for power grid facilities is improved by iteratively optimizing the model parameters. The trained model is evaluated using the validation set. The model's precision, recall rate, F1 score and other indicators are calculated, and the model prediction results are visualized. The model is then tuned based on the evaluation results.

[0033] The real-time remote sensing image is input into the trained YOLO target detection model. The model performs forward propagation on the image, identifies the power grid facilities and marks their locations and bounding boxes. Subsequently, the power grid facility locations and bounding box information predicted by the model are integrated with the original remote sensing image. The location of the power grid facilities can be marked by drawing a bounding box on the image or adding a label. The final remote sensing image marked with the power grid facility location and bounding box is the output of the model, which contains the location information and bounding box markings of the power grid facilities.

[0034] S220. Select sample points representing soil areas and non-soil areas in the remote sensing image, use the collected sample points to train a support vector machine, use the trained classifier to perform classification prediction on the remote sensing image, and obtain classification results of soil areas and non-soil areas.

[0035] When training a support vector machine (SVM), the features of each pixel are extracted from the remote sensing image, such as color information, texture features, shape features, etc. Feature extraction methods such as gray-level co-occurrence matrix (GLCM), local binary pattern (LBP), and color space conversion can be used to annotate the image data and mark it as soil areas and non-soil areas. The radial basis function kernel (RBF Kernel) is selected as the kernel function. Methods such as cross-validation can be used to select the best parameter combination to tune the parameters of the model. The data is divided into a training set and a test set for training and evaluating the performance of the model. Ultimately, the model can classify new remote sensing images into soil areas and non-soil areas, thereby obtaining classification results of soil areas and non-soil areas.

[0036] S300. Delineate a designated range area on the real-time remote sensing image, which is a soil area and is centered on the location of the power grid facility.

[0037] Taking into account that the impact of vegetation on power grid facilities is closely related to factors such as its distance from the power grid facilities and its growth degree, by setting an area of ​​specified size around the location of the power grid facilities, the vegetation area that may have an impact around the power grid facilities can be efficiently defined. The size of the area can be adjusted according to actual conditions. When judging the impact of vegetation on power grid facilities, since only a limited area near the power grid facilities needs to be analyzed instead of the entire remote sensing image, the amount of calculation can be significantly reduced, the efficiency of the algorithm can be improved, and the potential impact of vegetation on power grid facilities can be effectively identified and evaluated.

[0038] A designated area is defined as a soil area centered on the location of the power grid facility and specifically includes:

[0039] S310. Select a circular area with a radius of R0+k*I(s) centered on the power grid facility.

[0040] S320. Select a soil area within the circular area to obtain a designated range area; wherein R0 represents a basic radius, and k represents an adjustment coefficient for adjusting the radius size according to the degree to which the power grid facilities are susceptible to vegetation. Where α and β are parameters that control the shape of the function, s represents the type of power grid facilities, I(s)∈(0,1), I(s) is close to 1, indicating that the power grid facilities are easily affected by vegetation, and I(s) is close to 0, indicating that the power grid facilities are not easily affected by vegetation.

[0041] Taking the power grid facility as the center, a basic circular area with a radius of is selected. According to the susceptibility of the power grid facility to vegetation and the type of power grid facility, different radii are selected to more accurately evaluate the impact range of vegetation on the power grid facility. Different types of power grid facilities have different sensitivities to vegetation. For example, high-voltage lines are more susceptible to vegetation than low-voltage lines. Substation equipment may also be affected by vegetation blocking or touching. High-voltage lines require a larger impact range to ensure a safe distance, while low-voltage lines can be set with a smaller impact range. By dynamically adjusting the radius, introducing parameters for the shape of the control function and the type of power grid facility characteristics, and using the S-shaped curve function (Sigmoid function) to represent the degree of impact of power grid facilities on vegetation, the calculation of the adjustment coefficient k can be further refined, and the range of different power grid facilities affected by vegetation can be more accurately defined, making the evaluation results more accurate.

[0042] S400. Input the remote sensing image into the trained prediction model to obtain the predicted vegetation growth distribution within the specified future time and specified range.

[0043] The range of vegetation around power grid facilities is usually relatively limited, so only a detailed prediction of this area is needed to meet the needs. Narrowing the prediction range to the vegetation area around power grid facilities can significantly reduce the amount of data that the model needs to process, thereby greatly reducing the computational complexity and time cost, and can effectively improve the prediction speed and meet the needs of power grid management for rapid response.

[0044] S410. The remote sensing image has an infrared band and a visible light band. NDVI calculation is performed on the infrared band and the red band in the remote sensing image to obtain the NDVI value of each pixel.

[0045] The specific method is to use the radiation brightness values ​​of the infrared band (usually in the near-infrared band) and the red band for calculation. The calculation formula of NDVI is: In the formula, NIR represents the radiance value of the infrared band, and Red represents the radiance value of the red band. By calculating the NDVI value of each pixel, the vegetation coverage information corresponding to the pixel can be obtained. Common remote sensing image processing software and programming libraries provide tools and functions for calculating NDVI, which can easily calculate and extract NDVI values. The value range of NDVI is between -1 and +1. Negative values ​​usually represent very sparse vegetation or water bodies, and values ​​close to zero represent areas without vegetation such as rocks, bare land or deserts. Positive values ​​represent different degrees of vegetation coverage, and the higher the value, the denser the vegetation coverage.

[0046] S420. Input the remote sensing image with NDVI value into the trained prediction model, and output the predicted remote sensing image with NDVI value within the specified future time and specified range.

[0047] According to the designated range area delineated by S300, the remote sensing image is cropped, the information in the area is retained, the cropped image is data enhanced and the NDVI value is extracted, the NDVI values ​​of the historical remote sensing image are arranged in chronological order to form a time series, and the NDVI values ​​in the cropped remote sensing image are used as spatial features, which can be represented as a two-dimensional matrix. The convolutional recurrent neural network (CRNN) is trained to let the model learn to predict the distribution of NDVI values ​​in the designated range area in the future designated time period from the historical NDVI value sequence and spatial features, and the trained CRNN model is used to predict the distribution of NDVI values ​​in the designated range area in the future designated time period.

[0048] S430. Based on the real-time remote sensing image and the predicted remote sensing image with NDVI value, the predicted vegetation growth distribution within the specified time and range in the future is obtained.

[0049] Real-time remote sensing images provide information on the current vegetation type and growth level, while the predicted NDVI value images reflect the future trend of vegetation growth. Combining the two, combining NDVI values ​​with vegetation types and growth levels, can obtain more accurate and detailed vegetation growth distribution.

[0050] Based on real-time remote sensing images and predicted remote sensing images with NDVI values, the predicted vegetation growth distribution within a specified time and range in the future is obtained, including:

[0051] S431. Define the vegetation value VR of each pixel in the real-time remote sensing image ij , VR ij = {vr ij}, where v ij Represents the vegetation of the jth growth level of the i-th type in the real-time remote sensing image;

[0052] By defining the vegetation value VR of each pixel in the real-time remote sensing image ij , can refine vegetation information to the level of different types and growth degrees, so as to more accurately describe the real-time vegetation distribution, help capture the diversity and changes of vegetation, and enable the prediction model to more finely analyze and predict the distribution of vegetation of different types and growth degrees within a specified time and range. In addition, it also provides a more specific data basis for subsequent vegetation growth distribution analysis, making subsequent decision-making and planning more accurate and targeted, and enabling a more comprehensive understanding and utilization of vegetation information in remote sensing images.

[0053] S432. Calculate the vegetation value VP of each pixel in the remote sensing image with NDVI value within the specified future time and specified range based on the NDVI value. ij , VP ij for: NDVI pk ={vp ij}; where NDVI rk Represents the NDVI value of the kth pixel in the real-time remote sensing image, NDVI pk Represents the predicted NDVI value of the kth pixel in the remote sensing image, vp ij Represents the predicted vegetation of the jth growth level of the ith type in the remote sensing image.

[0054] By using NDVI as an indicator of vegetation growth and combining the vegetation values ​​in remote sensing images with NDVI values, we can more comprehensively reflect the growth of vegetation and more finely analyze the distribution of vegetation of different types and growth levels at a specified time in the future. Through detailed prediction of vegetation distribution, we can more accurately identify the potential impact of vegetation on power grid facilities, including the obstruction of transmission lines by different vegetation and the impact of vegetation on the stability and safety of facilities. This helps to promptly identify problems that may be caused by vegetation growth and provide more comprehensive information support for the planning, maintenance and management of power grid facilities.

[0055] S500. Evaluate the impact of predicted vegetation growth distribution on power grid facilities.

[0056] By demarcating a designated range area centered on the location of the power grid facility and the soil area, the vegetation growth distribution at a designated time in the future is predicted, and the impact of the predicted vegetation growth distribution on the power grid facilities is evaluated. This can quickly and accurately predict the vegetation growth distribution in the designated time and range in the future, thereby improving the efficiency of power grid facility management, reducing the impact of vegetation growth on power grid facilities, and ensuring the safety and stability of power grid operation.

[0057] Vegetation types include trees, shrubs, and grass.

[0058] Tall trees with dense branches and leaves can easily block the line of sight, making line maintenance more difficult. At the same time, their well-developed root systems may damage the foundations of power poles and affect the stability of facilities. Shrubs grow relatively short, but their dense branches and vines may also entangle power lines, causing short circuits or failures. Grasslands grow lower and have less direct impact on power grid facilities.

[0059] The impact of the predicted vegetation growth distribution on power grid facilities is evaluated in detail including:

[0060] Traverse each power grid facility to determine whether the power grid facility is covered by vegetation.

[0061] When power grid facilities are covered by vegetation, the degree of impact on power grid facilities is c For: Impac c =(w t V t +w s V s +w g V g )×C×F1; where V t 、V s and V g Represents the growth degree of trees, shrubs and grass respectively, w t 、w s and w gThey represent the weight coefficients of the impact of trees, shrubs and grasslands on power grid facilities, C represents the coverage of vegetation on power grid facilities, and F1 represents the comprehensive factor.

[0062] When the power grid facilities are not covered by vegetation, the degree of impact on the power grid facilities is d For: Impac d =(w t V t +w s V s +w g V g )×D×F2; where D represents the closest distance to the power grid facilities and F2 represents the comprehensive factor.

[0063] The impact of vegetation on power grid facilities is comprehensively considered, and the degree of impact is quantified into specific values, providing a scientific basis for subsequent evaluation and decision-making. By traversing and judging each power grid facility, we can fully understand the impact of vegetation on the facility, conduct targeted evaluation and management, consider the growth degree of different types of vegetation and the degree of impact on the facility, and introduce weight coefficients to make the evaluation more accurate and detailed. The introduction of comprehensive factors takes into account other influencing factors, making the evaluation more comprehensive and objective. Different impact degree calculation methods are given for the two situations of power grid facilities covered by vegetation and not covered by vegetation, making the evaluation more flexible and accurate, helping power grid managers to better understand the impact of vegetation on facilities, take targeted measures, and improve the safety and reliability of facility operation.

[0064] The coverage degree C of vegetation on power grid facilities is: Where N P Represents the number of pixels occupied by power grid facilities, N V Represents the number of pixels occupied by vegetation among the pixels occupied by power grid facilities.

[0065] Using the number of pixel points for comparison can concretize the vegetation coverage into numbers, avoiding the uncertainty caused by subjective judgment, improving the objectivity and accuracy of the evaluation, and does not require complex measurement equipment and techniques, so it can be easily applied in actual work.

[0066] Embodiment 2

[0067] This embodiment provides a vegetation growth model system for a power grid engineering scenario, which adopts the vegetation growth model method for a power grid engineering scenario as in Embodiment 1, and includes: an image acquisition module, a feature recognition module, a range selection module, a vegetation prediction module and an impact assessment module.

[0068] The image acquisition module is used to acquire remote sensing images of power grid engineering scenes, including real-time remote sensing images and historical remote sensing images; the feature recognition module is used to identify the location of power grid facilities and soil areas in real-time remote sensing images; the range selection module is used to delineate a specified range area centered on the location of the power grid facilities and the soil area on the real-time remote sensing image; the vegetation prediction module is used to input the remote sensing image into the trained prediction model to obtain the predicted vegetation growth distribution within the specified future time and specified range area; the impact assessment module is used to assess the impact of the predicted vegetation growth distribution on the power grid facilities.

[0069] This embodiment has all the advantages of the vegetation growth model method for power grid engineering scenarios in Embodiment 1, and can automatically implement all steps of the vegetation growth model method for power grid engineering scenarios.

[0070] Embodiment 3

[0071] This embodiment also provides a vegetation growth model device for a power grid engineering scenario, which includes a memory and a processor; the memory is used to store computer programs; the processor is used to implement the steps of the vegetation growth model method for a power grid engineering scenario in Example 1 when executing the computer program.

[0072] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps of the functions specified in a box or multiple boxes. Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they know the basic creative concept. Therefore, the attached claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

Claims

1. A vegetation growth model method for power grid engineering scenarios, characterized in that: It includes the following steps: Acquire remote sensing images of power grid engineering scenes, including real-time remote sensing images and historical remote sensing images; identifying power grid facility locations and soil regions in the real-time remote sensing image; Delineating a designated range area on the real-time remote sensing image, which is a soil area and is centered on the location of the power grid facility; Inputting the remote sensing image into a trained prediction model to obtain the predicted vegetation growth distribution within a specified future time and range; An impact of the predicted vegetation growth distribution on the power grid facilities is evaluated.

2. The vegetation growth model method for power grid engineering scenarios according to claim 1, characterized in that: A YOLO target detection model is trained using a remote sensing image with marked power grid facilities, and the real-time remote sensing image is input into the YOLO target detection model to obtain a remote sensing image with marked power grid facility locations and bounding boxes.

3. The vegetation growth model method for power grid engineering scenarios according to claim 1, characterized in that: Sample points representing soil areas and non-soil areas are selected in the remote sensing image, the collected sample points are used to train the support vector machine, and the trained classifier is used to classify and predict the remote sensing image to obtain the classification results of soil areas and non-soil areas.

4. The vegetation growth model method for power grid engineering scenarios according to claim 1, characterized in that: A designated area is defined as a soil area centered on the location of the power grid facility and specifically includes: Select a circular area with a radius of R0+k*I(s) centered on the power grid facility; Selecting a soil area within the circular area to obtain a designated range area; Among them, R0 represents the basic radius, k represents the adjustment coefficient for adjusting the radius size according to the susceptibility of power grid facilities to vegetation. Where α and β are parameters that control the shape of the function, s represents the type of power grid facilities, I(s)∈(0,1), I(s) is close to 1, indicating that the power grid facilities are easily affected by vegetation, and I(s) is close to 0, indicating that the power grid facilities are not easily affected by vegetation.

5. The vegetation growth model method for power grid engineering scenarios according to claim 2, characterized in that: The remote sensing image has an infrared band and a visible light band, and NDVI calculation is performed on the infrared band and the red band in the remote sensing image to obtain the NDVI value of each pixel; The remote sensing image with NDVI value is input into the trained prediction model, and the remote sensing image with NDVI value within the specified future time and specified range is output. Based on real-time remote sensing images and predicted remote sensing images with NDVI values, the predicted vegetation growth distribution within a specified time and range in the future can be obtained.

6. The vegetation growth model method for power grid engineering scenarios according to claim 5, characterized in that: Based on real-time remote sensing images and predicted remote sensing images with NDVI values, the predicted vegetation growth distribution within a specified time and range in the future is obtained, including: Define the vegetation value VR for each pixel in the real-time remote sensing image ij , VR ij = {vr ij }, where v ij Represents the vegetation of the jth growth level of the i-th type in the real-time remote sensing image; Calculate the vegetation value VP of each pixel in the remote sensing image with NDVI value within the specified future time and range based on the NDVI value ij , VP ij for: In the formula, NDVI rk Represents the NDVI value of the kth pixel in the real-time remote sensing image, NDVI pk Represents the predicted NDVI value of the kth pixel in the remote sensing image, vp ij Represents the predicted vegetation of the jth growth level of the ith type in the remote sensing image.

7. The vegetation growth model method for power grid engineering scenarios according to claim 6, characterized in that: Vegetation types include: trees, shrubs and grass.

8. The vegetation growth model method for power grid engineering scenarios according to claim 7, characterized in that: Evaluating the impact of the predicted vegetation growth distribution on the power grid facilities specifically includes: Traverse each power grid facility to determine whether it is covered by vegetation. When the power grid facilities are covered by vegetation, the degree of impact on the power grid facilities is c for: Impact c =(in t V t +in s V s +in g V g )×C×F1 Where V t 、V s and V g Represents the growth degree of trees, shrubs and grass respectively, w t 、w s and w g They represent the weight coefficients of the impact of trees, shrubs and grasslands on power grid facilities, C represents the coverage of vegetation on power grid facilities, and F1 represents the comprehensive factor; When the power grid facilities are not covered by vegetation, the degree of impact on the power grid facilities is d for: Impact d =(in t V t +in s V s +in g V g )×D×F2 In the formula, D represents the shortest distance to the power grid facilities, and F2 represents the comprehensive factor.

9. The vegetation growth model method for power grid engineering scenarios according to claim 7, characterized in that: The coverage degree C of vegetation on power grid facilities is: Where N P Represents the number of pixels occupied by power grid facilities, N V Represents the number of pixels occupied by vegetation among the pixels occupied by power grid facilities.

10. A vegetation growth model system for power grid engineering scenarios, which uses the vegetation growth model method for power grid engineering scenarios as described in any one of claims 1 to 9, characterized in that: It includes: An image acquisition module, which is used to acquire remote sensing images of power grid engineering scenes, including real-time remote sensing images and historical remote sensing images; A feature recognition module for identifying power grid facility locations and soil areas in the real-time remote sensing image; A range selection module, which is used to delineate a designated range area centered on the power grid facility location and being a soil area on the real-time remote sensing image; A vegetation prediction module, which is used to input the remote sensing image into a trained prediction model to obtain the predicted vegetation growth distribution within a specified time and range in the future; An impact assessment module is used to assess the impact of the predicted vegetation growth distribution on the power grid facilities.