Inverse Problem for Linear Systems Containing Uncertain Parameters

The concept of fuzzy dynamic programming and of guaranteed cost control initiated by Chang is applied to the inverse problem of guaranteed cost control for linear systems containing uncertain parameters. Obtained, for a class of linear systems with uncertain parameters, is a frequency-domain criterion which, analogous to that for the deterministic system, is sufficient for the control law to be of guaranteed cost type. The author presents an iterative method which, if convergent, enables one to solve the inverse problem for the general linear systems.