ICNPAA 2010: Mathematical Problems in
Engineering, Aerospace and Sciences

Neural Network based model predictive controller (NN-MPC) scheme for designing autonomous flight control of a rotary-wing unmanned aerial vehicle is presented in this paper. The applicability of the NN-MPC scheme is evaluated on a simplified heave model of the helicopter in simulation. Neural Network (NN) based system identification (NNID) technique is used to model the nonlinear dynamics of the unmanned helicopter. The NN model is then used in the MPC algorithm to estimate the future control moves. To show the efficacy of the NN-MPC scheme, controller results are provided. Results indicate that NN-MPC scheme is capable of handling external disturbances and parameter variations of the system.