Research On Measurement Method Of Model Surface Temperature And Ablation Quantity Under High Temperature Flow Field

In order to observe the ablation phenomenon of hypersonic vehicle head model under high temperature flow field and extract the surface temperature and ablation of hypersonic vehicle head model in whole ablation process.Firstly,based on the charge coupled device camera colorimetric temperature measurement mathematical model,A method for measuring the surface temperature of ablation model in high temperature flow field environment based on colorimetric temperature measurement is proposed,this method has the advantage of not measuring the emissivity in advance in a narrow band;Then based on the modified Canny edge extraction algorithm,A measurement method of ablation recession based on dynamic tracking of model edge points is designed,compared with the traditional ablation measurement method,it can directly show the whole ablation process of the model.The main research work of this paper is as follows:(1)Based on the theory of radiation temperature measurement and the working principle of CCD,the mathematical method of colorimetric temperature measurement for ablation model surface temperature measurement in high temperature flow field is established,and designed the video image acquisition system for model ablation process,The image acquisition system can effectively capture the ablation video image of the model in the temperature about 2000℃.(2)Because of the influence of noise from CCD camera and image acquisition process,the non local mean filtering algorithm is used to filter the ablation image noise.The image segmentation algorithm based on color information is used to segment the ablation image,the influence of background interference light on subsequent temperature calculation is eliminated.Spectral response coefficient of CCD camera obtained by calibration experiment,the temperature data of the ablation model surface are calculated and the temperature field distribution of the ablation model surface is extracted.Based on the temperature data collected by colorimetric pyrometer,the spectral emissivity change factor is calculated.(3)The existing Canny edge extraction algorithm is improved to extract the model edge in ablation image and the optimized treatment the edge image,the ablation recession data of edge points are extracted by edge tracking and edge tracking video image of the whole ablation back process is obtained.Combined with the extracted ablation recession data,the whole ablation recession process of feature edge points on the model surface is analyzed,based on the basic theory of recurrent neural network,the linear fitting prediction model of ablation quantity based on LSTM network is established,(4)According to the characteristics of nonlinear and over time of the ablation data,at the same time,the least square method is used to fit and predict the ablation data,the regression performance indexes of the two methods are evaluated,the experimental results show that the prediction effect of recurrent neural network is better.