Aerodynamic Moment Parameter Identification Based On QAR Data

The establishment of a precise aerodynamic moment model is important for the design of flight control system and it's also helpful in evaluation of flying qualities and flight simulation. Traditional aerodynamic parameters are identified through data from flight test or wind tunnel test with high cost. In this thesis, use the flight data recorded by the B737-800 aircraft Quick Access Recorder(QAR) to complete the parameter identification of aerodynamic moment model. The main work of this paper is as follows:First of all, establish the aerodynamic moment model of B737-800 aircraft. After the basic analysis of the attitude angular motion of the aircraft, set up the angular motion equations according to the principle of dynamics. Decouple the angular motion of the aircraft, and build the mathematical model of longitudinal motion and lateral motion separately.Secondly, preprocess the QAR data. Remove and correct the outliers in QAR data, obtain the angular velocity data and the posture angular acceleration data through differential smoothing method, improve the sampling rate of the original data by method of compressed sensing, and filter noise in different signals. On these bases, analyze the possibility of the parameter identification.Finally, identify the aerodynamic moment parameters. According to the principle of system identification, design the experiment for the entire identification process. Then, use least square method and maximum likelihood method to identify the model parameters of aerodynamic moment, analyze the identification results and verify the correctness of the identified parameters.Simulation results show that using QAR data for the aerodynamic moment model parameter identification, with QAR data preprocessed, can identify the various aerodynamic parameters correctly. The accuracy of the identification can meet the requirements of the simulation analysis and the results can be applied to flight simulation.