Research On Knowledge-Aided Detection Technology Of High Altitude Ice Crystals

Dangerous weather detection is very important to aviation safety.High altitude ice crystals are one of the weather phenomena that pose a great threat to flight safety.When an aircraft passes through an area with high ice crystals,the ice crystals will block the surface of the engine blades,causing the engine to stall and surge,and cause the combustion chamber to stall.In addition,ice crystals will enter the atmospheric data sensor,making the relevant sensor data abnormal,causing the aircraft to lose control.At present,the airborne weather radar installed on civil aviation aircraft is a single-polarization system,which can not effectively distinguish the type of precipitation from severe weather containing ice crystals.Therefore,it is of great value to carry out the research of high-altitude ice crystal detection technology.Aiming at the high-altitude ice crystal detection technology problem of airborne weather radar under the current single-polarization system,the thesis proposes a complete high-altitude ice crystal detection process based on the analysis of the ice phase particle composition,and uses auxiliary knowledge such as temperature and altitude to achieve effective detection of high-altitude ice crystals.The specific work of this thesis is as follows:Firstly,the composition of ice particles is analyzed,and then the calculation method of ice water content based on meteorological microphysical parameters is given.On this basis,a complete set of high-altitude ice crystal detection process is proposed.Secondly,the meteorological microphysical parameters of WRF model are used as the data source of ice crystal detection,and the height and temperature are analyzed.According to the auxiliary knowledge of temperature and height,the ice crystal area to be measured is selected,and then the radar data of the area to be measured is preprocessed.Then,according to the power law distribution function of ice crystal particle mass and the particle number concentration distribution function,the relationship between ice water content and radar reflectivity is modeled.Aiming at the exponential distribution of the general empirical formula,the proposed model is simplified,and the result shows that the model The distribution is consistent with the empirical formula.Finally,a high-altitude ice crystal detection method based on knowledge assistance and genetic algorithm is proposed.This method uses the height and temperature range of the ice crystal distribution area as auxiliary knowledge,predicts the existence of ice crystals in the airspace,and then performs quality control on the radar data obtained in the airspace.Secondly,using the prior values of ice crystal microphysical parameters as auxiliary knowledge,the 6 parameters to be estimated in the relationship model between ice water content and radar reflectivity are further simplified into 3 parameters to be estimated,and then the genetic algorithm is used to estimate the parameters.The specific operation It is to first generate an initial population for the three unknown parameters in the above model,and perform encoding,selection and mutation operations,then select the optimal solution according to the fitness function,and obtain the optimal parameters after decoding,and then obtain the determined ice water content and Radar reflectivity relationship model,and finally use this model to estimate the ice water content to achieve high-altitude ice crystal detection.Based on the analysis of simulation data of a real weather scene in Oklahoma,the experimental results show that the method can effectively estimate the ice water content and divide the ice crystal danger level,thereby effectively completing the high-altitude ice crystal detection.