Research On The Measurement Technology Of Air Data In Transonic And Hypersonic Flight

Air data system (ADS) is an essential airborne avionics system of aircraft. The measurement accuracy of air data is influenced by flight conditions and atmospheric environment.Mainly aimed at the key issues of violent fluctuation of air data in transonic flight, and air data computation model/algorithm in hypersonic flight for aircrafts, deep research concerning relevant air data measurement technology is carried out in traditional ADS and flush air data system (FADS) in the thesis.Aiming at altitude fluctuation of traditional ADS in transonic flight, the calculation principle of INS (Inertial Navigation System) altitude and barometric altitude is analyzed, and the characteristics of ADS/INS altitude-damp loop are studied. An ADS/INS fusion algorithm of transonic altitude is designed. The algorithm uses ADS/INS information with multiple fusion. The flight data of actual aircraft is applied to demonstrate the algorithm performance. The algorithm can reduce fluctuation amplitude of transonic ADS/INS altitude effectively, and improve the stationarity of altitude and flight stability.On the basis of analysis of airborne angle of attack (AOA) sensor configuration trait,calibration method and transonic AOA error characteristics, an transonic AOA compensation-correction algorithm based on BP neural network is designed for traditional ADS in the thesis. The neural network is trained and tested with improved BP algorithm and flight data. After verification and analysis, the compensation-correction algorithm can remarkably eliminate the violent jitter of transonic AOA, and enhance AOA accuracy and flight quality in transonic flight.The hypersonic aircraft (such as waverider vehicle) has faster speed and measures air data with FADS. In the thesis, three-dimensional geometric model of the waverider vehicle is constructed, and CFD(computational fluid dynamics) computing of the model is carried out in multiple flight conditions. The pressure field and temperature field distribution character of particular parts on the vehicle are analyzed, and the layout of FADS pressure-taps is designed and optimized with the CFD computing data.Based on CFD computing results of the waverider vehicle, the further model characteristic of FADS pressure field is analyzed. FADS algorithms are designed based on neural network. The model of pressure field is approached with BP network,improved BP network and RBF network separately. The algorithms are tested, and the results show improved BP network and RBF network can effectively solve Mach number and AOA/AOS(angle of sideslip) respectively. It is favorable to get accurate air data for the waverider vehicle.