When is a controller \(H^{\infty}\)-optimal ? (Q1109726): Difference between revisions

This paper gives conditions under which a given closed-loop, single- output, linear time-invariant control system is H-infinity optimal with respect to weighted combinations of its sensitivity and its complementary sensitivity functions. Here the weighting functions are assumed to be the outer factors of the inverse sensitivity and inverse complementary sensitivity functions respectively. The optimality criteria are simple formulas involving counting of zeros and poles of the plant, compensator and an auxiliary polynomial, and give new qualitative information for some simple examples. The main tool is a more general result of Helton which uses a winding number criterion to test when a given function is optimal for a very general class of H-infinity optimization problems.