An unmanned aerial vehicle route and high-altitude cableway wind field monitoring point optimization method

Abstract

The application discloses an unmanned aerial vehicle route and high-altitude cableway wind field monitoring point optimization method, and belongs to the technical field of wind engineering. The application obtains basic data through field measurement and numerical simulation, forms a standardized data set after preprocessing, builds an artificial intelligence model, constructs an evaluation index system including inversion accuracy, data representation ability and data stability, trains the model to comprehensively score the global potential point, and preliminarily screens out candidate points. Then, the preliminary point arrangement scheme is formed by adjusting the real constraints such as terrain, equipment installation, cost, operation and maintenance, and safety. Finally, the scheme is verified and corrected by using the existing monitoring data until the preset standard is met. The application realizes the scientific transformation of the point arrangement from artificial experience to data driving, guarantees the wind field inversion accuracy, and takes into account the engineering economy and practicability, thereby providing reliable wind field data support for the safe operation of the unmanned aerial vehicle route and the safe operation of the high-altitude cableway.

Description

Technical Field

[0001] This invention relates to the field of wind engineering technology, and in particular to a method for optimizing the monitoring points of wind fields using UAV flight paths and high-altitude cableways. Background Technology

[0002] In the selection and safe operation of low-altitude air routes and cable car routes, it is necessary to obtain wind field information in the corresponding areas in real time to support route planning and operational safety. However, wind field information along low-altitude air routes and cable car routes is highly dynamic and complex, making it difficult to capture directly and accurately. It often relies on the inverse calculation of the target wind field using surrounding measurement devices.

[0003] Currently, the deployment of surrounding measurement devices generally faces two major problems: First, some measurement points have low correlation with the main target wind field, making it difficult to accurately invert the wind field characteristics of the target area, resulting in insufficient reference value of the monitoring data; second, the deployment of measurement devices is limited by cost budgets, making it impossible to achieve uniform deployment across the entire area, and a balance must be struck between economy and monitoring effectiveness. Currently, the selection of measurement points largely relies on manual experience, which is highly subjective and lacks a scientific tool for quantifying correlation. This makes it difficult to accurately identify monitoring points with optimal prediction and inversion capabilities for the target wind field, thus failing to meet the dual requirements of accurate monitoring and cost control. Summary of the Invention

[0004] The purpose of this invention is to provide an optimization method for monitoring measurement points of UAV flight routes and high-altitude cableways, which solves the problems of blind layout of measurement points, insufficient inversion accuracy, and disconnection from actual engineering needs in existing wind field monitoring.

[0005] To achieve the above objectives, this invention provides a method for optimizing monitoring points for UAV flight paths and high-altitude cableway wind fields, comprising the following steps: S1. Preliminary survey and requirements definition of the monitoring site: Survey the terrain and obstacle information of the monitoring site, define the monitoring scope, and clarify the wind field monitoring indicators, accuracy and deployment constraints.

[0006] S2. Basic data classification and collection: On-site measured data are collected through temporary monitoring equipment and low-altitude detection by drones, and professional software is used to build models to generate numerical simulation data of the wind field across the entire site.

[0007] S3. Basic data preprocessing: The measured data and numerical simulation data collected in step S2 are subjected to outlier removal, missing value imputation, standardization, normalization and consistency testing to form a standardized dataset.

[0008] S4. Artificial Intelligence Algorithm Model Building and Training: Build an adapted data model, construct a point inversion performance evaluation index system, and train and optimize the model using a standardized dataset.

[0009] S5. Preliminary screening of potential monitoring points: Input the potential monitoring points across the entire region into the trained AI model, and select a set of candidate points with excellent inversion performance based on the comprehensive performance score.

[0010] S6. Optimize and adjust candidate locations based on real-world constraints: Based on the real-world constraints of terrain, equipment installation, cost, operation and maintenance, and safety, screen and adjust candidate locations to form a preliminary measurement point layout plan.

[0011] S7. Verification of Point Selection Results: Set up equipment at the initial measurement point locations to collect data, compare it with the measured data of existing measurement points, and verify whether the measurement point inversion accuracy meets the preset standards.

[0012] S8. Scheme Revision, Improvement and Final Determination: Adjust and revise the measurement points that fail the verification, re-verify until they meet the standards, and finally determine the optimal measurement point layout scheme.

[0013] Preferably, in step S4, the performance evaluation index system for point inversion is constructed, which specifically includes the following steps: Determine the data complexity of the standardized dataset; When the data is high-dimensional wind field data with multiple parameters, multiple locations, and multiple time periods coupled, a convolutional neural network model is selected; when the data is structured features corresponding to each potential monitoring location, a random forest regression model is selected. To construct a sample, for each potential monitoring point, extract its corresponding wind field parameters, spatial location parameters, and environmental parameters to form an input sample. The model outputs inversion accuracy score, data representation ability score and data stability score for each potential monitoring point, and calculates the comprehensive performance score through a weighted method; The formula for calculating the overall performance score is: ; in, For comprehensive performance scoring, To score the accuracy of the inversion, To score the ability to represent data, Score the data stability.

[0014] Preferably, the data representation ability score is quantified using the correlation coefficient, and its calculation formula is as follows: ; in, The correlation coefficient is... For candidate points in the th Wind field characteristic values ​​at each moment, For the target wind field in the first Reference value at a given moment. The average value of the candidate point sequence. The average value of the target wind field sequence, The total number of sample time points; Data stability scores are quantified using the coefficient of variation, and the calculation formula is as follows: ; in, The coefficient of variation is 1. The standard deviation of the continuously monitored data. This represents the average value of continuously monitored data.

[0015] Preferably, in step S4, when a random forest regression model is used, the model's output format is as follows: ; in, For the model prediction results, For the number of decision trees, For the first Each decision tree has input features The predicted output.

[0016] Preferably, in step S4, when a convolutional neural network model is used, the convolutional layer operation of the model is represented as follows: ; in, For the first Layer The output of each feature map This is the input feature map for the previous layer. For bias terms, For activation function, This is the convolution operator.

[0017] Preferably, in step S3, the data is standardized, and the expression is as follows: ; in, These are the original eigenvalues. These are the standardized eigenvalues. This is the minimum value of the feature in the sample. This represents the maximum value of the feature in the sample.

[0018] Preferably, in step S1, the site topography and surrounding obstacle distribution information are collected by combining drone aerial photography and on-site reconnaissance, and a site topography map with an accuracy of not less than 1:500 is drawn; the monitoring range is defined as: the entire drone flight path and the key influence area within 500m, or the entire high-altitude cableway and the key influence area within 500m.

[0019] Preferably, in step S2, the on-site measured data are collected using an ultrasonic anemometer temporarily deployed at a grid spacing of 50-100m, with a continuous collection time of no less than 72 hours; the numerical simulation data are generated by constructing a three-dimensional wind field numerical model using FLUENT or CFD software, with the simulation range covering the monitoring site and the surrounding 1km area.

[0020] Preferably, the practical constraints in step S6 include: terrain constraints, equipment installation constraints, cost constraints, operation and maintenance constraints, and safety constraints; among which, the equipment installation constraints require that the distance between the measuring point and the cableway support cable be no less than 5m, and the distance between the measuring point and the UAV take-off and landing point be no less than 10m.

[0021] Preferably, the verification criteria in step S7 are: the deviation between the inverted measurement data and the existing measured data of the measurement points is ≤5%, the correlation coefficient is ≥0.9, and the root mean square error is ≤0.2m / s.

[0022] Therefore, the method for optimizing UAV flight paths and high-altitude cableway wind field monitoring points using the above-described structure has the following beneficial effects: (1) This invention quantifies the inversion accuracy, data representation ability and stability of each point for the target wind field through artificial intelligence model, and initially screens out the points with the best inversion effect, which solves the problem of blind layout of measuring points and reliance on human experience, and realizes the transformation from subjective experience judgment to objective data-driven.

[0023] (2) This invention takes into account the practical constraints such as terrain conditions, equipment installation feasibility, monitoring costs, and operation and maintenance convenience, optimizes and adjusts the candidate points, and finally determines the optimal measurement point layout scheme that takes into account both scientificity and practicality, thus achieving the best balance between monitoring accuracy and engineering economy.

[0024] (3) This invention uses existing measured data to verify and correct the optimization scheme, forming a closed-loop optimization process, which further improves the accuracy and reliability of wind field inversion, and provides high-quality wind field data support for safe operation of UAV routes, safe operation of high-altitude cableways and risk prevention and control.

[0025] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the model construction process for an optimization method of monitoring and measuring points for wind field of UAV flight path and high-altitude cableway according to the present invention; Figure 2 This is a schematic diagram of a random forest structure for an optimization method of monitoring and measuring points for wind field monitoring of unmanned aerial vehicles and high-altitude cableways according to the present invention. Figure 3This is a schematic diagram of the convolutional neural network structure of the method for optimizing the monitoring points of UAV flight paths and high-altitude cableway wind fields according to the present invention. Detailed Implementation

[0027] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0028] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0029] Example like Figure 1 As shown, this invention provides a method for optimizing monitoring points for UAV flight paths and high-altitude cableway wind fields, including the following specific implementation steps: Step S1: Preliminary site survey and requirement definition for monitoring: Conduct a comprehensive preliminary survey of the monitoring site to clarify the core requirements and boundary conditions for wind field monitoring. First, using a combination of drone aerial photography and on-site reconnaissance, collect detailed information on the site's topography, the distribution, height, and dimensions of surrounding obstacles (buildings, mountains, trees, power lines, etc.), and create a site topographic map (accuracy no less than 1:500). Second, define the monitoring scope, precisely delineating the entire drone flight path area (including take-off and landing points, operating routes, and alternate landing areas), the entire high-altitude cableway (including the supporting cables, supports, and the area around the station), and key surrounding impact areas (within 500m on both sides of the flight path / cableway). Finally, based on actual engineering needs, clarify the core indicators for wind field monitoring (wind speed, wind direction, turbulence intensity, gust coefficient, etc.), monitoring accuracy requirements (wind speed accuracy ±0.1m / s, wind direction accuracy ±1°), monitoring duration, and core constraints on the layout of measuring points (such as not affecting the normal operation of the drone / cableway, equipment installation difficulty, etc.), and generate a site survey report and requirements specification.

[0030] Step S2: Basic data classification and collection: Based on the survey results and requirements of step S1, two types of core basic data were collected.

[0031] S21. On-site data collection: A dual data collection method of "temporary monitoring equipment + UAV low-altitude detection" is adopted. The temporary monitoring equipment uses ultrasonic anemometers, deployed at a grid spacing of 50-100m, prioritizing deployment at different elevations and terrains (flat areas, slopes, valleys, etc.) and areas prone to wind disturbance within the site, continuously collecting data for no less than 72 hours (covering different time periods during the day and night, and encompassing different meteorological conditions such as clear skies, light winds, and gusts), recording core parameters such as wind speed, wind direction, turbulence intensity, and turbulence integral scale in real time; the UAV uses a multi-rotor model equipped with miniature meteorological sensors, and conducts round-trip detection along a preset route (covering the entire monitoring area, with the flight altitude consistent with the UAV's operating route and the height of the high-altitude cableway's supporting cable), collecting a set of wind field data every 10m, and simultaneously recording the coordinates and elevation information of the collection points.

[0032] S22. Numerical Simulation Data Generation: Combining site topographic data, surrounding obstacle parameters, and regional meteorological background data for the past 5 years (average wind speed, wind direction frequency, extreme wind conditions), a three-dimensional wind field numerical model is constructed using professional fluid dynamics simulation software such as FLUENT or CFD. Reasonable boundary conditions are set (inlet wind speed is set according to the region's annual average wind speed, outlet pressure is standard atmospheric pressure, and ground roughness is adjusted according to the site's surface type). The simulation step size is set to 0.5s, and the simulation range covers the monitoring site and a surrounding 1km area. Wind field simulation data (including wind speed, wind direction distribution, wind field disturbance coefficients, etc.) are generated for all points across the site at different times, and standardized data files are exported.

[0033] Step S3, Basic Data Preprocessing: The measured and numerical simulation data collected in step S2 are systematically preprocessed. First, outliers are removed using the 3σ criterion (data exceeding the mean ± 3 standard deviations are considered outliers). For data collection points with a high number of outliers, data are recollected. Second, missing data are filled using linear interpolation (the missing rate is controlled within 5%; if the missing rate is too high, data is recollected). Then, all data are standardized and normalized (to eliminate the influence of dimensions). The expression for standardization is: ; in, These are the original eigenvalues. These are the standardized eigenvalues. This is the minimum value of the feature in the sample. This represents the maximum value of the feature in the sample.

[0034] Finally, a data consistency check is performed, comparing the deviations between the measured data and the simulated data to ensure that the deviation does not exceed 5%. After passing the check, the data is integrated to form a standardized dataset.

[0035] Step S4: Building and training the artificial intelligence algorithm model: A dedicated artificial intelligence algorithm model was built, and parameters were optimized through model training. First, the data complexity of the standardized dataset was determined. When the data mainly consisted of high-dimensional wind field data coupled with multiple parameters, multiple locations, and multiple time periods, and could be organized into regular grid data, two-dimensional matrix data, or three-dimensional tensor data according to spatial location and temporal order, a convolutional neural network model was selected. When the data mainly consisted of structured features such as wind speed, wind direction, turbulence intensity, gust coefficient, turbulence integral scale, location coordinates, location elevation, and spatial relationship parameters corresponding to each potential monitoring point, and the samples could be organized into location feature vectors, a random forest regression model was selected.

[0036] After determining the model type, samples were constructed from the standardized dataset. For each potential monitoring point, its corresponding wind field parameters, spatial location parameters, and environmental parameters were extracted to form the input samples. The model does not directly output a binary judgment result regarding whether to deploy a monitoring point; instead, it first outputs the inversion accuracy score for each potential monitoring point. Data representation ability score and data stability score Then, calculate the overall performance score. The overall performance score is calculated using a weighted method, and the formula is as follows: ; Among them, the data representation ability score P_{rep} is preferably quantified using the correlation coefficient r, and the expression is: ; in, The correlation coefficient is... For candidate points in the th Wind field characteristic values ​​at each moment, For the target wind field in the first Reference value at a given moment. The average value of the candidate point sequence. The average value of the target wind field sequence, This represents the total number of sample times.

[0037] Data stability score The coefficient of variation is preferred. Quantization is performed, and the expression is: ; in, The coefficient of variation is 1. The standard deviation of the continuously monitored data. This represents the average value of continuously monitored data.

[0038] When using a random forest regression model, such as Figure 2 As shown, the model mainly consists of a data input module, a feature processing module, a decision tree ensemble module, and a comprehensive scoring output module. The output of the random forest can be expressed as: ; in, For the model prediction results, For the number of decision trees, For the first Each decision tree has input features The predicted output.

[0039] When using a convolutional neural network model, such as Figure 3 As shown, the model mainly consists of a data input module, a convolutional feature extraction module, a pooling module, a fully connected module, and a scoring output module. The basic operations of the convolutional layer can be represented as: ; in, For the first Layer The output of each feature map This is the input feature map for the previous layer. For bias terms, For activation function, This is the convolution operator.

[0040] During the model training phase, the standardized dataset is divided into training and test sets in a 7:3 ratio for training and optimization of the model parameters. After training, the model's prediction accuracy is validated using the test set to ensure that the model's prediction accuracy is not less than 90%. Once the preset standard is met, the model is saved.

[0041] Step S5: Preliminary screening of potential locations: Using a trained artificial intelligence model, potential monitoring points are divided into 10m×10m grids, covering the entire drone flight path, the entire aerial cableway, and surrounding key areas. Relevant data from all potential points are input into the trained model, which quantifies and scores each point based on three key evaluation indicators, calculating a comprehensive performance score (out of 100) according to weighted criteria. A comprehensive performance score threshold (e.g., 85 points) is set based on engineering requirements, and points with a comprehensive score ≥ 85 points and all individual indicators meeting or exceeding the "good" standard are selected to form a candidate point set.

[0042] Step S6: Optimize and adjust candidate points based on real-world constraints: The shortlisted candidate sites were optimized and adjusted based on the actual constraints of the engineering site, with a focus on: ① Terrain constraints: avoiding areas with complex terrain, slopes greater than 30°, or those prone to landslides or mudslides; ② Equipment installation constraints: ensuring that equipment installation does not affect the operation of the UAV flight path or the normal operation of the aerial cableway (distance from the cableway's supporting cable not less than 5m, and distance from the UAV's take-off and landing point not less than 10m); ③ Cost constraints: prioritizing sites where existing installation foundations can be reused; ④ Operation and maintenance constraints: selecting sites with good accessibility and convenient equipment maintenance; ⑤ Safety constraints: avoiding high-voltage power lines and flammable and explosive areas. Through comprehensive consideration, sites that did not meet the constraints were eliminated, and backup sites were added, forming a preliminary survey point layout plan.

[0043] Step S7: Verification of point selection results: Temporary equipment is deployed at each initially set measurement point to continuously collect wind field data for no less than 48 hours, which serves as the measurement point inversion data. Measured data from existing measurement points during the same period are retrieved, and the deviation, correlation coefficient, and root mean square error (RMSE) are calculated. Verification standards are set as follows: deviation ≤ 5%, correlation coefficient ≥ 0.9, and RMS error ≤ 0.2 m / s. If all measurement points meet the verification standards, the preliminary scheme is considered successful; if some measurement points do not meet the standards, the reasons for the deviation are recorded, and the process proceeds to the next correction stage.

[0044] Step S8: Scheme Revision, Improvement, and Finalization: The issues identified during verification were corrected as follows: if the inversion deviation was caused by the terrain of the measurement point, the point was replaced with a backup point that had similar performance and better terrain conditions; if the deviation was due to equipment installation errors, the installation angle and height of the equipment were adjusted, and data was collected again for verification; if the number of measurement points was insufficient, high-performance measurement points were added. After correction, verification was performed again until all measurement points met the verification criteria. Finally, detailed information on all measurement points was compiled, and a measurement point layout diagram and equipment installation instructions were prepared to form the final optimized measurement point layout plan.

[0045] Therefore, the present invention adopts the above-mentioned method for optimizing the monitoring points of UAV flight routes and high-altitude cableways, which can achieve precise optimization of the deployment of monitoring points and provide reliable wind field data support for the safe operation of UAV flight routes and the safe operation of high-altitude cableways.

[0046] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for optimizing monitoring points for UAV flight paths and high-altitude cableway wind fields, characterized in that, Includes the following steps: S1. Preliminary survey and requirements definition of the monitoring site: Survey the topography and obstacle information of the monitoring site, define the monitoring scope, and clarify the wind field monitoring indicators, accuracy and deployment constraints; S2. Basic data classification and collection: On-site measured data are collected through temporary monitoring equipment and low-altitude detection by drones, and professional software is used to build models to generate numerical simulation data of the wind field in the entire site. S3. Basic data preprocessing: The measured data and numerical simulation data collected in step S2 are subjected to outlier removal, missing value imputation, standardization, normalization and consistency testing to form a standardized dataset; S4. Artificial Intelligence Algorithm Model Building and Training: Build an adapted data model, construct a point inversion performance evaluation index system, and train and optimize the model using a standardized dataset. S5. Preliminary screening of potential monitoring points: Input the potential monitoring points across the entire region into the trained AI model, and select a set of candidate points with excellent inversion performance based on the comprehensive performance score. S6. Optimize and adjust candidate locations based on real-world constraints: Based on the real-world constraints of terrain, equipment installation, cost, operation and maintenance, and safety, screen and adjust candidate locations to form a preliminary measurement point layout plan; S7. Verification of point selection results: Set up equipment at the initial measurement point locations to collect data, compare it with the measured data of existing measurement points, and verify whether the measurement point inversion accuracy meets the preset standard. S8. Scheme Revision, Improvement and Final Determination: Adjust and revise the measurement points that fail the verification, re-verify until they meet the standards, and finally determine the optimal measurement point layout scheme.

2. The method for optimizing monitoring points for UAV flight paths and high-altitude cableway wind fields according to claim 1, characterized in that: In step S4, a performance evaluation index system for point inversion is constructed, which specifically includes the following steps: Determine the data complexity of the standardized dataset; When the data is high-dimensional wind field data with multiple parameters, multiple locations, and multiple time periods coupled, a convolutional neural network model is selected; when the data is structured features corresponding to each potential monitoring location, a random forest regression model is selected. To construct a sample, for each potential monitoring point, extract its corresponding wind field parameters, spatial location parameters, and environmental parameters to form an input sample. The model outputs inversion accuracy score, data representation ability score and data stability score for each potential monitoring point, and calculates the comprehensive performance score through a weighted method; The formula for calculating the overall performance score is: ； in, For comprehensive performance scoring, To score the accuracy of the inversion, To score the ability to represent data, Score the data stability.

3. The method for optimizing UAV flight paths and high-altitude cableway wind field monitoring points according to claim 2, characterized in that: Data representation ability is quantified using the correlation coefficient, and its calculation formula is as follows: ； in, The correlation coefficient is... For candidate points in the th Wind field characteristic values ​​at each moment, For the target wind field in the first Reference value at a given moment. The average value of the candidate point sequence. The average value of the target wind field sequence, The total number of sample time points; Data stability scores are quantified using the coefficient of variation, and the calculation formula is as follows: ； in, The coefficient of variation is 1. The standard deviation of the continuously monitored data. This represents the average value of continuously monitored data.

4. The method for optimizing UAV flight paths and high-altitude cableway wind field monitoring points according to claim 2, characterized in that: In step S4, when using the random forest regression model, the model's output format is as follows: ； in, For the model prediction results, For the number of decision trees, For the first Each decision tree has input features The predicted output.

5. The method for optimizing UAV flight paths and high-altitude cableway wind field monitoring points according to claim 2, characterized in that: In step S4, when a convolutional neural network model is used, the convolutional layer operation of the model is represented as follows: ； in, For the first Layer The output of each feature map This is the input feature map for the previous layer. For bias terms, For activation function, This is the convolution operator.

6. The method for optimizing monitoring points for UAV flight paths and high-altitude cableway wind fields according to claim 1, characterized in that: In step S3, the data is standardized, and the expression is as follows: ； in, These are the original eigenvalues. These are the standardized eigenvalues. This is the minimum value of the feature in the sample. This represents the maximum value of the feature in the sample.

7. The method for optimizing monitoring points for UAV flight paths and high-altitude cableway wind fields according to claim 1, characterized in that: In step S1, the site topography and surrounding obstacle distribution information are collected by combining drone aerial photography and on-site reconnaissance, and a site topography map with an accuracy of no less than 1:500 is drawn; the monitoring scope is defined as: the entire drone flight path and the key influence area within 500m, or the entire high-altitude cableway and the key influence area within 500m.

8. The method for optimizing monitoring points for UAV flight paths and high-altitude cableway wind fields according to claim 1, characterized in that: In step S2, the field measured data are collected using an ultrasonic anemometer temporarily deployed at a grid spacing of 50-100m, with a continuous collection time of no less than 72 hours; the numerical simulation data are generated by constructing a three-dimensional wind field numerical model using FLUENT or CFD software, with the simulation range covering the monitoring site and the surrounding 1km area.

9. The method for optimizing monitoring points for UAV flight paths and high-altitude cableway wind fields according to claim 1, characterized in that: The practical constraints in step S6 include: terrain constraints, equipment installation constraints, cost constraints, operation and maintenance constraints, and safety constraints; among them, the equipment installation constraints require that the distance between the measuring point and the cableway support cable be no less than 5m, and the distance between the measuring point and the UAV take-off and landing point be no less than 10m.

10. The method for optimizing monitoring points for UAV flight paths and high-altitude cableway wind fields according to claim 1, characterized in that: The verification criteria in step S7 are: the deviation between the inverted data of the measuring point and the existing measured data of the measuring point is ≤5%, the correlation coefficient is ≥0.9, and the root mean square error is ≤0.2m / s.

Images