| Civil Aviation flight attitude of the aircraft, flight stability, flight path are all affected by aerodynamic. The development of the aerodynamic force model can not only help develop aircraft, but also provide essential premise for simulation devices such as flight simulator and maintenance simulator. At present, aerodynamic force parameter identification based on flight test data has done well in parameter process, mathematical model and identification methods, and this way is better than identification with wind-tunnel test data, which can only get aerodynamic parameters in special moment. However, the flight test data is usually unavailable because of its confidentiality. Choose the data recorded by the airborne Quick Access Recorder(QAR) to analyze aerodynamic force and identify aerodynamic force parameters can reflect the influence of flight environment on the aircraft, and save the research cost at the same time.Build the aerodynamics equations after analysis on aircraft movement model and mechanics model. Use the data partition method to cut down the bad influence caused by nonlinear factors such as alpha and mach. Establish the linear model set. Through research and analysis on QAR data, process the data in order to remove errors and abnormal values led by measurement and decrease the noise caused by high frequency. Transform the airspeed in QAR data into true airspeed after study on their relationship. Then, according to data partition theory, part the data set into several segments on basis of the nonlinear influences of different range of alpha and mach values. After studying on system identification principles and current identification methods, analyzing the features of QAR data and aerodynamic force mathematical model, choose the majorizing algorithm of least square method based on genetic algorithm for parameter identification. During the process of simulation experiment, solve the problem of data selection and partition, as well as the problem of thrust simplified calculation. Modify the affect on kinetic pressure brought by crosswind, choose the data for genetic algorithm, and finally get a good identification result. |
