The Research Of Modeling Based On System Identification And Validation For Tandem Ducted Fan Land And Air Vehicle

Land and air vehicle can be applied widely in both military and civilian areas as a kind of integrated transportation tool which is able to run on the land and fly in the air. In order to obtain the excellent control performance for the land and air vehicle, a mathematical model that accurately describes the vehicle’s dynamics is always required.However, the high nonlinearities, cross-couplings and dynamic uncertainties of land and air vehicle bring about the main difficulties in mechanism modeling especially when the researchers don’t have enough experience about the aerodynamics. In this paper, the modeling of the scale tandem ducted fan land and air vehicle model combining the mechanism and system identification experiments was studied.Firstly, this paper analyzed the drive system models including the brushless DC motor(BLDC), and the nonlinearities physical aerodynamics modeling of the full-vehicle was also presented including the body components modeling such as ducted fan. Then the nonlinearities physical aerodynamics model was linearized to the linear state space model.Secondly, based on the frequency-domain identification theory the identified experimental schemes were designed, and the corresponding software were programed to collect the experimental data.Thirdly, the drive system models including the brushless DC motor and servo motor were obtained through the identified experiments, and the time-domain validation experiments for the drive system identified models also be doneFinally, the flight tests for identifying angular dynamics models of land and air vehicle were carried out. Based on the collected tests data and the mechanism model, the SISO transfer functions for simplified each channel for MIMO angular dynamics state space model were identified to estimate the initial values for unknown model parameters. Then the initial values were applied to identify the two-channel state-space models, and all the unknown aerodynamic and control derivatives of MIMO angular dynamics state space model were obtained. The time-domain validation showed the excellent agreement between the identified models with the flight tests data.