Font Size: a A A

Ice Thickness Prediction Of Airport Runway Area Based On Least Square Support Vector Machine

Posted on:2021-01-31Degree:MasterType:Thesis
Country:ChinaCandidate:J H LiuFull Text:PDF
GTID:2392330611968721Subject:Control engineering
Abstract/Summary:PDF Full Text Request
Aiming at the prediction of the ice thickness of Airport Road area,the heat balance model in the process of rainfall is established,and the ice growth law under the influence of air temperature,humidity,wind speed,rainfall and road surface material is analyzed.According to the orthogonal experimental table,the experimental scheme is designed,and the experimental data are used to predict the ice thickness of the road area by least square support vector regression.The main contents of this paper are as follows:Firstly,based on the energy conservation,the energy balance equation of ice accretion process is established to solve the freezing coefficient;the mass of ice accretion in the body is controlled according to the trajectory of water droplets;the collision coefficient of water droplets on the road surface is calculated by the Euler two-phase flow model.In this paper,the influence of meteorological factors on the ice growth is analyzed.Secondly,the prediction vector of ice accretion is composed of meteorological factors and ice thickness.Aiming at the problem of ice thickness of airport runway area,a prediction model of ice thickness of airport runway area based on the least square support vector machine(LS-SVM)parameter optimization by particle swarm optimization(PSO)algorithm is established.A new learning factor parameter optimization range and a linear decreasing inertia weight are introduced into PSO algorithm.The ice thickness of the road area is predicted according to the ice deposition prediction model.Thirdly,set up the road surface simulation experiment platform,simulate and control the change range of meteorological factors to carry out the experiment,according to the experimental results,get the primary and secondary order of influencing factors of Road area ice.The accuracy of the established ice growth model and pso-ls-svr prediction model is verified.The validity of the model is verified by real-time ice prediction experiment.
Keywords/Search Tags:airport pavement, support vector machine, least squares support vector machine, particle swarm optimization algorithm, ice thickness, orthogonal experiment
PDF Full Text Request
Related items