System Identification and Tracking Control of a Flapping Wing Micro Air Vehicle

Abstract

The goal of system identification is to build mathematical models of dynamical systems that can be used in control techniques. This thesis is focused on identifying the dynamics of a flapping wing micro air vehicle (FWMAV) from measured flight data. Because of the nonlinear and complex dynamics and coupled variables of a FWMAV, discrete-time linear time-invariant (LTI) models for both the longitudinal and lateral dynamics which have dominant flight state variables are used as a basis for system identification. To gather input-output data sets from flight test, an experimental setup is constructed in an indoor environment using a motion capture system. Discrete-time LTI models are sought by employing linear estimator, support vector regression (SVR) and validated in the time domain flight data. The results show that SVR accurately produces the true FWMAV responses better than the linear estimator. After dynamic characteristics of a FWMAV are checked through matrix analysis, the obtained model is used to design feedback controllers for maintaining altitude. To verify the controller, a simulation is performed. Experimental results, using a FWMAV, of maintaining a flight altitude are also presented.