Research On Analytical Method And Function Construction Method Of Aircraft Telemetry Data

With the development of science and technology,the production and application of aircraft are more and more extensive.In order to ensure the formal flight safety of aircraft,a lot of testing work needs to be done.The thesis study the analysis and function construction methods of the telemetry data to improve the fitting accuracy and applicability of the algorithm,The thesis use IRIG 106 standard complete frame integrity judgment,data shunting and data conversion,and constructs the function of the engine speed,airspeed,temperature and humidity data.This thesis studies the least squares method,five-point cubic smoothing algorithm,median filter and wayelet threshold algorithm,and designs a wavelet algorithm which improve the threshold function,and verifies the validity of the algorithm by MATLAB.The research contents are as follows:(1)The thesis uses C-H++-QT framework to complete the design of data monitoring software,realizes the configurability of telemetry engineering tasks and data analysis processing,which simplifies the hardware structure.The thesis uses the technology based on multi-buffer pool to realize the telemetry data frame shunting storage,which shunts the muiti-class parameters of the integration path,and improves the data analysis efficiency;(2)The function construction methods of least squares method,five-point cubic smoothing and median filtering are studied in this thesis,The results show that the least squares polynomial fitting method has slow convergence speed,when the polynomial order is small,it can,t achieve smooth effect,When the polynomial order is too large,it will appear over-fittmg phenomenon and increase noise removal error.The five-point cubic smoothing can effectively remove data noise,but there are serious data jump phenomenon,which can't achieve the effect of curve smoothing.The median filtering,whose smoothing window length is variable,and can better adjust the smoothing effect of curve display,but it has poor de-noising accuracy,which only pays attention to the trend of curve change,but ignores the authenticity of data;(3)The construction nethod of improving wavelet threshold function,the algorithm effectively improves the data jump and over-smoothing phenomena existing in the traditional wavelet threshold function.It can effectively deal with the various data noise of telemetry system,reduce the measurement error and make the measured value closer to the true value,Beside it,it has strong applicability and smoothness,which make the curve similarity is above 80%,and can reflect the original measurement more truthfully.The studies of data analysis and function construction methods meets the real-time requirement of system data analysis,which increase the portability and flexibility of the system and improves the defects of large measurement error and weak applicability in common algorithm to make aviation flight safety.