Lateral acceleration control design of a non-linear homing missile using multi-objective evolutionary optimisation

We present the lateral acceleration control design of nonlinear missile model using a multiobjective evolutionary optimisation method (NSGA-II -like). The controller design for the uncertain plants is carried out by minimising gain and phase margins and tracking performance objectives of the corresponding vertices. Pareto surfaces are used to identify a feasible control structure and analyse its performance tradeoffs. Based on the selected trade-off solution, the interpolated controller, whose poles, zeros and gains are linear continuous functions of Mach number and incidence, are designed for the whole operating envelope. The interpolated controller is now synthesised by minimising the Euclidean distance of multiple operating points' objective values. The stability is preserved by additionally overlapping these operating regions. The nonlinear simulation results show that the resulting interpolated controller is indeed a robust tracking controller for all possible perturbations.