Parameter optimization in systems subject to worst (bounded) disturbance

In a given system which is subject to the worst bounded disturbance that maximizes a chosen performance index, the best parameters can be so determined that the same performance index is minimized. The disturbance, which is assumed deterministic and bounded, acts only for some fixed time, known a priori. The Maximum Principle is applied in the determination of the worst disturbance [1]. An algorithm which yields the worst disturbance and the best parameters is developed by employing in the parameter optimization an approximation technique based on the steepest descent. Computational results are presented, and the existence of the solution is discussed.