CN-121766546-B - Unmanned aerial vehicle suspended cargo safety evaluation method based on low-altitude wind direction deflection prediction

CN121766546BCN 121766546 BCN121766546 BCN 121766546BCN-121766546-B

Abstract

The invention relates to an unmanned aerial vehicle suspended cargo safety assessment method based on low-altitude wind direction deflection prediction, and belongs to the technical field of unmanned aerial vehicle application. The method comprises the steps of collecting meteorological data of a suspension area, real-time attitude and sling state parameters of an unmanned aerial vehicle through sensors carried by the unmanned aerial vehicle, constructing a safety association data set, extracting a low-altitude wind direction deflection rule from the data through a cross entropy time sequence prediction algorithm, constructing a multi-dimensional safety evaluation model integrating wind direction deflection influence, sling state, unmanned aerial vehicle attitude and risk duration by combining with the dynamic characteristics of the unmanned aerial vehicle, training and optimizing model parameters through historical data of different suspension scenes, setting a multi-stage evaluation index, a safety threshold and influence weight of the multi-stage evaluation index, and finally calculating the comprehensive safety level of the unmanned aerial vehicle in the process of suspending cargoes based on the model and the index. The wind direction change risk can be predicted in advance, multidimensional dynamic safety assessment is realized, and safety accidents caused by air interference are effectively reduced.

Inventors

LAI Yuanwen
WANG SHUYI
CHEN BINGLING
XU LING
ZHANG LEI
HUANG XINTONG
SHEN LING
YANG YUCHUN
LI JIANQUAN
XU QIN

Assignees

福州大学

Dates

Publication Date: 20260512
Application Date: 20260303

Claims (8)

1. The unmanned aerial vehicle suspended cargo safety evaluation method based on low-altitude wind direction deflection prediction is characterized by comprising the following steps of: Acquiring low-altitude image data of a suspension area, real-time attitude parameters and sling state parameters of the unmanned aerial vehicle through a weather sensing module and a suspension state monitoring module carried by the unmanned aerial vehicle, preprocessing the acquired data, and constructing a safety association data set D; Based on the safety association data set D, predicting wind direction deflection by using a cross entropy time sequence prediction algorithm, and constructing a multi-dimensional suspended cargo safety evaluation model by combining with the mechanical characteristics of an unmanned aerial vehicle; Training the constructed multi-dimensional suspended cargo safety evaluation model by adopting historical safety association data sets under different suspension scenes, and adjusting model parameters through iterative optimization; setting a multi-level evaluation index system comprising the duration of risk and the swing angle of a sling, and determining the safety threshold value of each level of index and the influence weight of each level of index in a multi-dimensional suspended cargo safety evaluation model; based on the evaluation model after training and optimization and the set evaluation index, calculating a comprehensive safety evaluation value M in the process of suspending the cargo by the unmanned aerial vehicle, and dividing the safety grade according to the value interval of the M; the suspension area is an unmanned aerial vehicle suspension low-altitude airspace with a distance of 0-1000 meters from the ground, and the low-altitude image data of the suspension area comprises low-altitude instantaneous wind speed Instantaneous wind direction θ Intensity of atmospheric turbulence I Turbulent pulse frequency f The real-time attitude parameters of the unmanned aerial vehicle comprise the pitch angle alpha of the unmanned aerial vehicle Roll angle beta Yaw angle gamma Take-off and landing speed v uav The sling state parameters include sling swinging angle Angular velocity omega of sling swing After preprocessing the collected data, the constructed safety association data set D is expressed as: In the formula, Is the first The time of the sampling is the same as the time of the sampling, The total sampling times; Based on the safety association data set D, predicting the wind direction deflection by using a cross entropy time sequence prediction algorithm, and specifically comprises the following steps: Defining a standard wind direction for suspended cargo of an unmanned aerial vehicle Calculating real-time wind direction deflection The calculation formula is as follows: Wherein, the Is the first Real-time wind direction deflection quantity at moment, I ) Is the first The atmospheric turbulence intensity at the moment is 0.1 which is the atmospheric turbulence influence coefficient; real-time wind deflection by adopting cross entropy time sequence prediction algorithm Prediction is carried out, and a wind direction deflection prediction sequence { of a future T period is output T=10-30 seconds, the core prediction formula is: Wherein, the To the future (future) The predicted value of the wind direction deflection at the moment, To the point of Epsilon (tau) is a turbulence disturbance error term for the weight coefficient of the prediction model; Statistical prediction in sequence ≥ As the risk duration T r , in which And the wind direction deflection early warning threshold value is set.
2. The unmanned aerial vehicle suspended cargo safety evaluation method based on low-altitude wind direction deflection prediction according to claim 1, wherein the expression for constructing the multi-dimensional suspended cargo safety evaluation model by combining the mechanical characteristics of an unmanned aerial vehicle is as follows: Wherein M is a comprehensive safety evaluation value; Is a wind direction deflection influencing factor; is a sling state factor; Is a gesture stabilization factor; is the duration of the risk; 、 the weight coefficient of the wind direction deflection influence factor, the weight coefficient of the sling state factor, the weight coefficient of the attitude stabilization factor and the weight coefficient of the risk duration factor respectively meet the following requirements The value ranges are all 0.2-0.5.
3. The unmanned aerial vehicle suspended cargo safety assessment method based on low-altitude wind direction deflection prediction according to claim 2, wherein the wind direction deflection influencing factor The calculation formula of (2) is as follows: In the formula, For the future The average value of the wind direction deflection predicted values of the time period, Is the limit wind deflection threshold.
4. The unmanned aerial vehicle suspended cargo safety assessment method based on low-altitude wind direction deflection prediction according to claim 2, wherein the sling state factor The calculation formula of (2) is as follows: Wherein, the , , Is the standard swinging angle of the sling, Is the standard swing angular velocity of the sling, Is used for limiting the swing angle of the sling.
5. The unmanned aerial vehicle suspended cargo safety assessment method based on low-altitude wind direction deflection prediction according to claim 2, wherein the attitude stabilization factor The calculation formula of (2) is as follows: Wherein, the 、、 , For the extreme angular deviation of the attitude, And (5) suspending the cargo standard attitude angle for the unmanned aerial vehicle.
6. The unmanned aerial vehicle suspended cargo safety evaluation method based on low-altitude wind direction deflection prediction according to claim 2, wherein the constructed multi-dimensional suspended cargo safety evaluation model is trained by adopting historical safety association data sets under different suspension scenes, and model parameters are adjusted through iterative optimization, and the method specifically comprises the following steps: dividing a historical data set covering different hanging scenes of plain, mountain, city and sea into a training set and a verification set in proportion; Inputting the training set into a multi-dimensional suspended cargo safety evaluation model for iterative training, aiming at minimizing evaluation errors, and adjusting the weight coefficient of the prediction model by a gradient descent method To the point of Evaluating model weight coefficients 、 ; Stopping training when the evaluation error variation of the verification set of the continuous multiple iterations is smaller than a set threshold value or the iteration number reaches an upper limit; verifying whether the prediction accuracy of the trained model is more than or equal to 90 percent and the evaluation error is less than or equal to 5 percent, and readjusting the model parameters for training if the indexes are not met.
7. The unmanned aerial vehicle suspended cargo safety assessment method based on low-altitude wind direction deflection prediction according to any one of claims 2 to 6, wherein the weight coefficient is 、 Dynamically determining according to the real-time risk level of the corresponding factor, wherein: weighting factor of wind direction deflection influencing factor Determining according to the risk level of the wind direction deflection influence factor; Weight coefficient of sling state factor Determining according to the risk level of the sling state factor; Weight coefficient of attitude stabilization factor Determining according to the risk level of the attitude stabilization factors; Weight coefficient of risk duration factor And determining according to the risk level of the risk duration.
8. The unmanned aerial vehicle suspended cargo safety evaluation method based on low-altitude wind direction deflection prediction according to claim 1, wherein the safety classes are classified according to the comprehensive safety evaluation value M, wherein: when M < 0.3, the risk level is "safe"; when M is more than or equal to 0.3 and less than 0.6, the risk level is 'low risk'; when M is more than or equal to 0.6 and less than or equal to 0.8, the risk level is 'stroke risk'; When M is more than or equal to 0.8, the risk grade is 'high risk'.

Description

Unmanned aerial vehicle suspended cargo safety evaluation method based on low-altitude wind direction deflection prediction Technical Field The invention belongs to the technical field of unmanned aerial vehicle application, and particularly relates to an unmanned aerial vehicle suspended cargo safety assessment method based on low-altitude wind direction deflection prediction. Background Along with the rapid iteration of unmanned aerial vehicle technology, the application scene of the unmanned aerial vehicle has been extended from traditional aerial mapping, inspection and monitoring to the complex suspended cargo operation fields such as material transportation, engineering hoisting, emergency rescue, offshore wind power maintenance and the like. The unmanned aerial vehicle suspended cargo operation plays an irreplaceable role in the scenes of urban high-rise cargo transportation, mountain emergency cargo delivery, offshore platform equipment hoisting and the like by virtue of the technical advantages of flexibility, high efficiency and no limitation of topography. However, the high safety accident rate caused by air interference has become a core bottleneck for restricting the large-scale landing of the field. The low-altitude environment has the remarkable characteristics of variable wind speed and frequent turbulence, wherein the coupling effect of wind direction deflection and atmospheric turbulence can directly lead to the instability of the attitude of the unmanned aerial vehicle and the severe swing of a sling, so as to cause serious safety accidents such as falling of suspended goods, collision barriers, falling of the unmanned aerial vehicle and the like. For example, in the operation of hanging goods in urban high-rise buildings, the 'narrow pipe effect' among the high-rise buildings can lead to the sudden increase of 3-5 times of local wind speed, sudden wind direction deflection is extremely easy to cause the collision of slings and building outer walls so as to cause the damage or falling of suspended goods, in the offshore wind power maintenance scene, the irregular pulsation and turbulence interference of sea wind can cause the swing angle of slings to exceed 30 degrees, the falling of suspended goods is extremely easy to be caused, the safety of operators and equipment is seriously threatened, in the mountain emergency rescue scene, the wind direction of the mountain valley wind alternately day and night changes, the gesture is out of control when the unmanned aerial vehicle hangs the rescue goods, the goods delivery deviation or falling is caused, and the key rescue opportunity is delayed. The existing unmanned aerial vehicle suspended cargo safety evaluation method has obvious limitations that firstly static evaluation is carried out by relying on real-time meteorological data, prediction is not carried out on a low-altitude wind direction deflection rule, potential risks in the cargo suspending process are difficult to avoid in advance, secondly evaluation indexes are single and focused on unmanned aerial vehicle attitude or wind speed parameters, coupling influences of sling states, cargo swing amplitude and wind direction deflection are ignored, thirdly model generalization capability is insufficient, cargo suspending historical data of different operation scenes are not combined for training optimization, and evaluation accuracy is difficult to match actual operation requirements. Therefore, there is a need for an unmanned aerial vehicle suspended cargo safety assessment method capable of accurately predicting low-altitude wind direction deflection and taking cargo safety influence factors into consideration in a multi-dimensional manner, so as to break through the bottleneck of the prior art and provide effective technical support for the safe development of unmanned aerial vehicle suspended cargo operation. Disclosure of Invention The invention aims to solve the problems that the existing unmanned aerial vehicle suspended cargo safety evaluation lacks wind direction deflection prediction capability, has single evaluation index, has weak model generalization capability and the like, and provides an unmanned aerial vehicle suspended cargo safety evaluation method based on low-altitude wind direction deflection prediction, which is beneficial to accurately quantifying the safety risk of an unmanned aerial vehicle in the low-altitude suspended cargo process and improving the safety guarantee level of the unmanned aerial vehicle suspension. In order to achieve the purpose, the technical scheme of the invention is that the unmanned aerial vehicle suspended cargo safety assessment method based on low-altitude wind direction deflection prediction comprises the following steps: Acquiring low-altitude image data of a suspension area, real-time attitude parameters and sling state parameters of the unmanned aerial vehicle through a weather sensing module and a suspension state monitoring modul