Unmanned aerial vehicle taking-off and landing site selection method based on NHPP risk field and multi-dimensional comprehensive evaluation

Abstract

The invention discloses an unmanned aerial vehicle taking-off and landing site selection method based on an NHPP risk field and multi-dimensional comprehensive evaluation. The method comprises the following steps: S1, creating n1 candidate areas in a research area; s2, collecting space-time risk monitoring data of all candidate areas of the research area, inputting the space-time risk monitoring data into the space-time non-homogeneous Poisson process wind disaster risk model, and calculating to obtain NHPP risk cumulative strength corresponding to the candidate areas; s3, constructing a multi-dimensional comprehensive evaluation system including P1 dimensions including NHPP risk cumulative strength and construction cost, and performing weighted evaluation to obtain an evaluation value of the candidate region; and selecting an optimal candidate area as a take-off and landing site selection area, or setting a decision threshold value and outputting a site selection decision result according to the decision threshold value. According to the method, the NHPP risk fields of all candidate areas are constructed through the space-time inhomogeneous Poisson process wind disaster risk model, the space-time data are effectively utilized to construct the NHPP risk fields, the candidate areas of the unmanned aerial vehicle take-off and landing field are scientifically and reasonably evaluated in combination with multiple dimensions, and reliable technical support is provided for low-altitude aviation site selection decision.

Description

Technical Field

[0001] The present invention relates to the field of low-altitude aviation decision-making technology, and in particular to a method for selecting a take-off and landing site for unmanned aerial vehicles (UAVs) based on a NHPP risk field and multi-dimensional comprehensive evaluation. Background Art

[0002] In the field of low-altitude aviation, infrastructure and intelligent decision-making are key to its development. The rational assessment and site selection of drone landing sites presents a technical challenge that needs to be addressed. U.S. Patent No. 11741702B2 discloses a method for automatically selecting safe landing sites for drone systems. The method proposes generating a depth map by acquiring two overlapping images, identifying flat areas, and calculating a "landing area quality score" based on depth variance and semantic type. The area with the highest score is selected as the landing site. Canadian Patent No. CA2977945A1 discloses an environment scanning and tracking drone system. This system uses LiDAR and cameras to scan candidate areas in real time, constructing an environmental map, and then scoring and screening areas that meet flatness and obstacle thresholds. Currently, single-factor risk or threshold alerts are also used. Some existing technologies only monitor wind speed or obstacle density and decide whether to enable or suspend landings based on fixed thresholds. These methods lack temporal and spatial aggregation of risks and multi-dimensional trade-off analysis. These three methods rely solely on one-way monitoring or periodic manual review, failing to accurately and scientifically assess drone landing site selection. Summary of the Invention

[0003] The purpose of the present invention is to provide a method for UAV take-off and landing site selection based on NHPP risk field and multi-dimensional comprehensive evaluation. The NHPP risk field of all candidate areas is constructed through the spatiotemporal non-homogeneous Poisson process wind disaster risk model. The spatiotemporal NHPP risk field is applied to the take-off and landing site selection assessment for the first time. A multi-dimensional comprehensive evaluation system is used for weighted evaluation and the optimal candidate area is selected as the take-off and landing site selection area. The spatiotemporal data is effectively used to construct the NHPP risk field and the multi-dimensional scientific and reasonable evaluation of the candidate areas of the UAV take-off and landing site is combined, providing reliable technical support for low-altitude aviation site selection decision-making.

[0004] The purpose of the present invention is achieved through the following technical solutions:

[0005] A method for selecting a takeoff and landing site for unmanned aerial vehicles (UAVs) based on NHPP risk fields and multi-dimensional comprehensive assessments, comprising:

[0006] S1, create n1 candidate regions in the study area;

[0007] S2. Collect spatiotemporal risk monitoring data from all candidate areas in the study area and input them into the constructed spatiotemporal non-homogeneous Poisson process wind disaster risk model to construct the NHPP risk field in spatiotemporal dimensions, and calculate the cumulative intensity of NHPP risk corresponding to the candidate areas;

[0008] S3. Construct a multi-dimensional comprehensive evaluation system that includes P1 dimensions, including the cumulative intensity of NHPP risks and construction costs, and perform weighted evaluation to obtain the evaluation value of the candidate area; select the optimal candidate area as the take-off and landing site selection area, or set a decision threshold and output the site selection decision result according to the decision threshold.

[0009] To better implement the present invention, if the study area is non-mountainous, the spatiotemporal risk monitoring data includes historical wind monitoring spatiotemporal series data, and the spatiotemporal non-homogeneous Poisson process wind disaster risk model constructs the NHPP risk field according to the spatiotemporal intensity function. The spatiotemporal intensity function expression is as follows:

[0010] ,in For the spatial position ,time The expected frequency of the event, For historical events to spatial locations ,time The risk impact of For the additional contribution to the risk when the wind speed exceeds the safety threshold, the historical events are the wind monitoring data recording points.

[0011] Preferably, if the study area is a mountainous area, the spatiotemporal risk monitoring data includes historical wind monitoring spatiotemporal series data, and the spatiotemporal non-homogeneous Poisson process wind disaster risk model constructs the NHPP risk field according to the risk density function. The risk density function expression is as follows:

[0012] ,in The spatial location of the mountainous area ,time The risk density, is the mountain event point i at time location, is the weight of historical event point i, is the wind speed at time t, is the safe wind speed threshold, is the spatial impact diffusion scale.

[0013] Preferably, the risk impact The expression is as follows:

[0014] ,in is the weight of historical event point i, is the location of historical event point i, is the decay rate of historical event point i with distance, is the risk impact coefficient when the wind speed exceeds the safety threshold;

[0015] The additional contribution The expression is as follows:

[0016] ,in is the risk impact coefficient when the wind speed exceeds the safety threshold, is the wind speed at time t, is the safe wind speed threshold.

[0017] Preferably, the candidate region The cumulative intensity of NHPP risk in the interval [t0,T] The expression is as follows: .

[0018] Preferably, the evaluation value of the candidate region It is obtained based on the weighted calculation of dimensional indicators in the multi-dimensional comprehensive evaluation system. The expression is as follows: ,in Dimensional indicators The assessed value of Dimensional indicators The weight of .

[0019] Preferably, the cumulative intensity of NHPP risk is verified and compared as follows: select sample locations and deploy weather stations, lidars and cameras at the sample locations through 5G cellular and LoRaWAN networks to monitor data and evaluate the observed risk. The predicted risk of the sample location is calculated by the spatiotemporal non-homogeneous Poisson process wind disaster risk model , will predict the risk and observation risk Difference .

[0020] Preferably, based on the difference Adjust the parameters of the spatiotemporal non-homogeneous Poisson process wind disaster risk model, and the expression is as follows:

[0021] ,in is the original parameter, is the adjusted parameter, is the risk sensitivity threshold, is the learning rate or scaling factor.

[0022] Preferably, if the study area is a mountainous area and mountain rescue is carried out, the multi-dimensional comprehensive evaluation system also includes two dimensions: communication quality evaluation and ground rescue accessibility evaluation. The communication quality evaluation and ground rescue accessibility evaluation are performed through drones to detect edge wind fields, detect air channels, and shoot videos, and the detection data are fed back to the command center in real time for evaluation.

[0023] Preferably, the evaluation value of the candidate area is a positive evaluation value, and the larger the evaluation value of the candidate area, the better it is; the site selection decision results include recommendation, conditional recommendation and non-recommendation.

[0024] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0025] (1) This paper constructs the NHPP risk field of all candidate areas through the spatiotemporal non-homogeneous Poisson process wind disaster risk model. It applies the spatiotemporal NHPP risk field to the site selection assessment of take-off and landing sites for the first time, uses a multi-dimensional comprehensive assessment system to perform weighted assessment and select the best candidate area as the take-off and landing site selection area. It effectively utilizes spatiotemporal data to construct the NHPP risk field and combines multi-dimensional scientific and reasonable assessment of the candidate areas of UAV take-off and landing sites, providing reliable technical support for low-altitude aviation site selection decisions and promoting the development of low-altitude intelligent decision-making technology.

[0026] (2) The present invention obtains a difference between the observed risk and the predicted risk of the sample point, and adjusts the parameters of the spatiotemporal non-homogeneous Poisson process wind disaster risk model based on the difference, realizing a closed-loop feedback and online correction parameter adjustment mechanism. A dynamic response mechanism is formed through group collaboration and task feedback loop, which significantly improves the accuracy of risk field assessment (error ≤ 5%) and adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 The figure is a flow chart of the method for selecting a site for take-off and landing of a UAV according to the present invention. DETAILED DESCRIPTION

[0028] Below in conjunction with embodiment, the present invention is described in further detail:

[0029] Example 1

[0030] like Figure 1 As shown, a method for selecting a site for UAV takeoff and landing based on NHPP risk field and multi-dimensional comprehensive assessment is provided, which includes:

[0031] S1. Create n1 candidate regions in the study area.

[0032] S2. Collect the spatiotemporal risk monitoring data of all candidate areas in the study area and input them into the constructed spatiotemporal non-homogeneous Poisson process wind disaster risk model to construct the NHPP risk field in the spatiotemporal dimension, and calculate the cumulative intensity of NHPP risk corresponding to the candidate areas.

[0033] If the study area is non-mountainous, the spatiotemporal risk monitoring data includes historical wind monitoring spatiotemporal series data. The spatiotemporal non-homogeneous Poisson process wind disaster risk model constructs the NHPP risk field according to the spatiotemporal intensity function. The spatiotemporal intensity function expression is as follows:

[0034] ,in For the spatial position ,time The expected frequency of the event, For historical events to spatial locations ,time The risk impact of the risk factor decreases exponentially with increasing distance; is the additional contribution to the risk when the wind speed exceeds the safety threshold, and the historical events are the wind monitoring data recording points. The expression is as follows:

[0035] ,in is the weight of historical event point i, is the location of historical event point i, is the decay rate of historical event point i with distance, is the risk factor when the wind speed exceeds the safety threshold. The expression is as follows:

[0036] ,in is the risk impact coefficient when the wind speed exceeds the safety threshold, is the wind speed at time t, is the safe wind speed threshold.

[0037] If the study area is a mountainous area, the spatiotemporal risk monitoring data includes historical wind monitoring spatiotemporal series data. The spatiotemporal non-homogeneous Poisson process wind disaster risk model constructs the NHPP risk field according to the risk density function. The risk density function expression is as follows:

[0038] ,in The spatial location of the mountainous area ,time The risk density, is the mountain event point i at time location, is the weight of historical event point i, is the wind speed at time t, is the safe wind speed threshold, is the spatial impact diffusion scale.

[0039] Candidate regions of the present invention The cumulative intensity of NHPP risk in the interval [t0,T] The expression is as follows: , where if the study area is non-mountainous area, for If the study area is a mountainous area, for The cumulative intensity of NHPP risk reflects the temporal and spatial aggregation trend and risk exposure.

[0040] In some embodiments, the cumulative intensity of NHPP risk is verified and compared as follows: select sample locations and deploy weather stations, lidars, and cameras at the sample locations through 5G cellular and LoRaWAN networks to monitor data and evaluate the observed risk. The predicted risk of the sample location is calculated by the spatiotemporal non-homogeneous Poisson process wind disaster risk model , will predict the risk and observation risk Difference (Calculate the difference between sample points to overestimate or underestimate the frequency of actual wind disasters). Adjust the parameters of the spatiotemporal non-homogeneous Poisson process wind disaster risk model, and the expression is as follows:

[0041] ,in is the original parameter, is the adjusted parameter, is the risk sensitivity threshold (to control the parameter update amplitude and avoid over-correction), is the learning rate or proportional factor (controls the correction speed and stability). Perform closed-loop feedback and online optimization (adjust parameters ), reflecting the risk sensitivity adjustment after feedback from observed data residuals. This paper uses differential adjustment to adjust the parameters of the spatiotemporal inhomogeneous Poisson process wind disaster risk model, implementing a closed-loop feedback and online correction and parameter adjustment mechanism. This dynamic response mechanism, formed through group collaboration and task feedback loops, significantly improves the accuracy of risk field assessment (error ≤ 5%) and its adaptability.

[0042] S3, build a multi-dimensional comprehensive evaluation system including P1 dimensions including NHPP risk accumulation intensity and construction cost, and perform weighted evaluation to obtain the evaluation value of the candidate area. It is obtained based on the weighted calculation of dimensional indicators in the multi-dimensional comprehensive evaluation system. The expression is as follows: ,in Dimensional indicators The assessed value of Dimensional indicators In some embodiments, if the study area is a mountainous area and mountain rescue is carried out, the multi-dimensional comprehensive evaluation system also includes two dimensions: communication quality evaluation and ground rescue accessibility evaluation. The communication quality evaluation and ground rescue accessibility evaluation are performed by drones to detect edge wind fields, detect air channels, and shoot videos, and the detection data is fed back to the command center in real time for evaluation.

[0043] The optimal candidate area is selected as the take-off and landing location, or a decision threshold is set and a location decision result is output according to the decision threshold. In some embodiments, the candidate area's evaluation value is a positive evaluation value, with a larger candidate area's evaluation value indicating a better candidate area. The location decision result includes three types: recommendation, conditional recommendation, and non-recommendation.

[0044] Example 2

[0045] A method for selecting a drone takeoff and landing site based on NHPP risk fields and multi-dimensional comprehensive assessments is described. This embodiment is used to select a landing site for a city logistics center. (City A wishes to establish a multi-rotor drone landing site near its southwest logistics center for high-frequency delivery operations. The logistics center address is "City A, Wanshou Road Logistics Park," which is located in a non-mountainous area.) The method includes:

[0046] S1. Create n1 candidate areas in the study area. Specifically, set a circular area with a radius of 1.5 km, centered around the Wanshou Road Logistics Park, and demarcate several candidate areas. Using high-resolution remote sensing imagery or map data (such as OpenStreetMap or urban construction BIM data), extract all open land, building rooftops, parking lots, and open squares within the circular area as candidate points. Apply a spatial clustering algorithm to merge adjacent or contiguous areas of the candidate points to form several candidate areas.

[0047] S2. Collect the spatiotemporal risk monitoring data of all candidate areas in the study area and input them into the constructed spatiotemporal non-homogeneous Poisson process wind disaster risk model to construct the NHPP risk field in the spatiotemporal dimension, and calculate the cumulative intensity of NHPP risk corresponding to the candidate areas.

[0048] Since Wanshou Road Logistics Park is a non-mountainous area, the spatiotemporal risk monitoring data includes historical wind monitoring spatiotemporal series data. The spatiotemporal non-homogeneous Poisson process wind disaster risk model constructs the NHPP risk field according to the spatiotemporal intensity function. The spatiotemporal intensity function expression is as follows:

[0049] ,in For the spatial position ,time the expected frequency of the event; is the weight of historical event point i, is the location of historical event point i, is the decay rate of historical event point i with distance, is the risk impact coefficient when the wind speed exceeds the safety threshold, is the wind speed at time t, is the safety wind speed threshold, in this embodiment It is 7m / s (the maximum safe take-off and landing wind speed for multi-rotors).

[0050] Candidate regions of the present invention The cumulative intensity of NHPP risk in the interval [t0, T] is verified and compared as follows: sample locations are selected and weather stations, lidars, and cameras are deployed at the sample locations through 5G cellular and LoRaWAN networks to monitor data and evaluate the observed risk. The predicted risk of the sample location is calculated by the spatiotemporal non-homogeneous Poisson process wind disaster risk model , will predict the risk and observation risk Difference (Calculate the difference between sample points to overestimate or underestimate the frequency of actual wind disasters). Adjust the parameters of the spatiotemporal non-homogeneous Poisson process wind disaster risk model, and the expression is as follows:

[0051] ,in is the original parameter, is the adjusted parameter, is the risk sensitivity threshold (to control the parameter update amplitude and avoid over-correction), is the learning rate or proportional factor (controls the correction speed and stability). Perform closed-loop feedback and online optimization (adjust parameters ), reflecting the risk sensitivity adjustment after the residual feedback of the observed data. The above method is an error-driven model parameter update method, which is a nonlinear residual feedback correction mechanism. The main logic is as follows: If the model predicts R pred Significant deviation from the observed R obs , that is, |ΔR| is large, which means that the risk response of some areas / time periods in the model is too high or too low; by enlarging or shrinking the relevant parameters To improve the fitting accuracy of the model; R th Ensure the same scale correction effect for risk fields of different scales; Control the correction "step size" to prevent oscillation or overfitting. For example: Assume that the model parameters of a certain area are , the model predicts risk R pred =0.6, but the observation is R obs =0.4, that is, ΔR=0.2. Let R th =1.0,κ=0.1, then , that is, slightly adjust the parameters in this area upwards to more accurately match the observed data.

[0052] S3. Construct a multi-dimensional comprehensive evaluation system including the cumulative intensity of NHPP risk and construction cost, including P1, and perform weighted evaluation to obtain the evaluation value of the candidate area. In the Wanshou Road Logistics Park as the study area, the multi-dimensional comprehensive evaluation system includes not only the cumulative intensity of NHPP risk and construction cost, but also the distance cost from the logistics center. , terrain characteristics (slope, obstruction) , land accessibility or usability score The evaluation value of the candidate area in this embodiment is Based on the multi-dimensional comprehensive evaluation system, the dimensional indicators are first normalized (the evaluation value of the candidate area is a positive evaluation value, and the larger the evaluation value, the better it is. During the normalization process, the larger the normalized dimensional indicator, the better it is. For example, the normalization formula for the cumulative intensity of NHPP risk is: , the distance cost normalization formula is: ), weighted processing is calculated, and the expression is as follows: ,in Dimensional indicators The assessed value of Dimensional indicators The best candidate area is selected as the take-off and landing site area.

[0053] Example 3

[0054] A method for selecting drone takeoff and landing sites based on NHPP risk fields and multi-dimensional comprehensive assessments. This embodiment is used for selecting drone takeoff and landing sites in mountainous emergency rescue scenarios. (During sudden disasters in mountainous areas, such as earthquakes, landslides, and mudslides, traditional ground transportation often fails due to road interruptions. Drones are playing an increasingly critical role in timely and flexible emergency rescue. However, complex mountainous terrain, unpredictable weather, and frequent signal blind spots pose significant challenges to drone takeoff and landing safety and mission scheduling. Therefore, selecting takeoff and landing sites with minimal risk and maximum response efficiency is a core technical challenge in emergency drone systems.) The method includes:

[0055] S1. Create n1 candidate regions in the study area. Take a typical mountainous area (e.g., an earthquake-prone mountainous area in Sichuan) as the study area, construct a 3D digital elevation model (DEM), and select several candidate regions.

[0056] S2. Collect the spatiotemporal risk monitoring data of all candidate areas in the study area and input them into the constructed spatiotemporal non-homogeneous Poisson process wind disaster risk model to construct the NHPP risk field in the spatiotemporal dimension, and calculate the cumulative intensity of NHPP risk corresponding to the candidate areas.

[0057] In this example, a mountainous area in Sichuan with frequent earthquakes is used as the study area. The spatiotemporal risk monitoring data includes historical spatiotemporal series data of wind monitoring. The spatiotemporal non-homogeneous Poisson process wind disaster risk model constructs the NHPP risk field according to the risk density function. The risk density function expression is as follows:

[0058] ,in The spatial location of the mountainous area ,time The risk density, is the mountain event point i at time location, is the weight of historical event point i, is the wind speed at time t, is the safe wind speed threshold, is the spatial impact diffusion scale.

[0059] Candidate regions of the present invention The cumulative intensity of NHPP risk in the interval [t0,T] The expression is as follows: , where if the study area is non-mountainous area, for If the study area is a mountainous area, for .

[0060] S3, build a multi-dimensional comprehensive evaluation system including P1 dimensions including NHPP risk accumulation intensity and construction cost, and perform weighted evaluation to obtain the evaluation value of the candidate area. It is obtained based on the weighted calculation of dimensional indicators in the multi-dimensional comprehensive evaluation system. The expression is as follows: ,in Dimensional indicators The assessed value of Dimensional indicators The study area of this embodiment is a mountainous area and mountain rescue (for simplicity, the construction cost is set to zero). The multi-dimensional comprehensive evaluation system also includes two dimensions: communication quality evaluation and ground rescue accessibility evaluation. The communication quality evaluation and ground rescue accessibility evaluation are performed by drones to detect edge wind fields, detect air channels, and shoot videos, and the detection data is fed back to the command center in real time for evaluation. The evaluation value of the candidate area The expression is: ,in is the communication quality evaluation result, This is the result of the ground rescue accessibility assessment. is the cumulative intensity of NHPP risk, are the weights of NHPP risk accumulation intensity, communication quality assessment, and ground rescue accessibility assessment, respectively. , for example, 0.5, 0.3, and 0.2, respectively. The optimal candidate area is selected as the takeoff and landing site selection area, or a decision threshold is set and the site selection decision result is output according to the decision threshold. In a mountainous area in Sichuan prone to earthquakes, three candidate sites were selected after evaluation using the method in this embodiment. Subsequent multi-aircraft surveys and verification ultimately led to the selection of a temporary landing pad located in an open area at the valley bottom (low risk, good signal quality, and accessible to traffic), which effectively supported emergency delivery and casualty evacuation missions.

[0061] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for selecting a site for UAV takeoff and landing based on NHPP risk fields and multi-dimensional comprehensive assessment, characterized by: The methods include: S1, create n1 candidate regions in the study area; S2. Collect spatiotemporal risk monitoring data from all candidate areas in the study area and input them into the constructed spatiotemporal non-homogeneous Poisson process wind disaster risk model to construct the NHPP risk field in spatiotemporal dimensions, and calculate the cumulative intensity of NHPP risk corresponding to the candidate areas; S3. Construct a multi-dimensional comprehensive evaluation system that includes P1 dimensions, including the cumulative intensity of NHPP risks and construction costs, and perform weighted evaluation to obtain the evaluation value of the candidate area; select the optimal candidate area as the take-off and landing site selection area, or set a decision threshold and output the site selection decision result according to the decision threshold.

2. The method for selecting a UAV takeoff and landing site based on NHPP risk field and multi-dimensional comprehensive assessment according to claim 1 is characterized by: If the study area is non-mountainous, the spatiotemporal risk monitoring data includes historical wind monitoring spatiotemporal series data. The spatiotemporal non-homogeneous Poisson process wind disaster risk model constructs the NHPP risk field according to the spatiotemporal intensity function. The spatiotemporal intensity function expression is as follows: ,in For the spatial position ,time The expected frequency of the event, For historical events to spatial locations ,time The risk impact of For the additional contribution to the risk when the wind speed exceeds the safety threshold, the historical events are the wind monitoring data recording points.

3. The method for selecting a UAV takeoff and landing site based on NHPP risk field and multi-dimensional comprehensive assessment according to claim 1 is characterized by: If the study area is a mountainous area, the spatiotemporal risk monitoring data includes historical wind monitoring spatiotemporal series data. The spatiotemporal non-homogeneous Poisson process wind disaster risk model constructs the NHPP risk field according to the risk density function. The risk density function expression is as follows: ,in The spatial location of the mountainous area ,time The risk density, is the mountain event point i at time location, is the weight of historical event point i, is the wind speed at time t, is the safe wind speed threshold, is the spatial impact diffusion scale.

4. The method for selecting a UAV takeoff and landing site based on NHPP risk field and multi-dimensional comprehensive assessment according to claim 2 is characterized by: The impact of the risks The expression is as follows: ,in is the weight of historical event point i, is the location of historical event point i, is the decay rate of historical event point i with distance, is the risk impact coefficient when the wind speed exceeds the safety threshold; The additional contribution The expression is as follows: ,in is the risk impact coefficient when the wind speed exceeds the safety threshold, is the wind speed at time t, is the safe wind speed threshold.

5. The method for selecting a UAV takeoff and landing site based on NHPP risk field and multi-dimensional comprehensive assessment according to claim 1 is characterized by: The candidate region The cumulative intensity of NHPP risk in the interval [t0,T] The expression is as follows: .

6. The method for selecting a UAV takeoff and landing site based on NHPP risk field and multi-dimensional comprehensive assessment according to claim 1 is characterized by: The evaluation value of the candidate region It is obtained based on the weighted calculation of dimensional indicators in the multi-dimensional comprehensive evaluation system. The expression is as follows: ,in Dimensional indicators The assessed value of Dimensional indicators The weight of .

7. The method for selecting a UAV takeoff and landing site based on NHPP risk field and multi-dimensional comprehensive assessment according to claim 1 or 5, characterized in that: The cumulative intensity of NHPP risk is verified and compared according to the following method: sample locations are selected and weather stations, lidars and cameras are deployed at the sample locations through 5G cellular and LoRaWAN networks to monitor data and evaluate the observed risk. The predicted risk of the sample location is calculated by the spatiotemporal non-homogeneous Poisson process wind disaster risk model , will predict the risk and observation risk Difference .

8. The method for selecting a UAV takeoff and landing site based on NHPP risk field and multi-dimensional comprehensive assessment according to claim 7 is characterized by: Based on the difference Adjust the parameters of the spatiotemporal non-homogeneous Poisson process wind disaster risk model, and the expression is as follows: ,in is the original parameter, is the adjusted parameter, is the risk sensitivity threshold, is the learning rate or scaling factor.

9. The method for selecting a UAV takeoff and landing site based on NHPP risk field and multi-dimensional comprehensive assessment according to claim 3 is characterized by: If the study area is a mountainous area and rescue is carried out in mountainous areas, the multi-dimensional comprehensive evaluation system also includes two dimensions: communication quality evaluation and ground rescue accessibility evaluation. The communication quality evaluation and ground rescue accessibility evaluation are carried out through drones to detect edge wind fields, detect aerial channels and shoot videos, and the detection data are fed back to the command center in real time for evaluation.

10. The method for selecting a UAV takeoff and landing site based on NHPP risk field and multi-dimensional comprehensive assessment according to claim 1, characterized in that: The evaluation value of the candidate area is a positive evaluation value, and the larger the evaluation value of the candidate area, the better it is; the site selection decision results include recommendation, conditional recommendation and non-recommendation.

Images