Optimal feedback control synthesis using simulated annealing techniques

This paper deals with the synthesis of robust controllers using the parameter optimization method. The controller synthesis process consists of two steps: 1) determination of a suitable controller structure and 2) optimization of controller parameters subjected to various engineering constraints. It is shown that heuristic optimization methods such as simulated annealing can be applied to significantly improve the parameter optimization process. A fifth order aircraft control problem is considered in this paper where it is demonstrated that a suitably optimized third order controller can produce closed loop performance comparable to those obtained by an 8th order LQG/LTR controller and an 18th order H_∞ controller