Model Error Sensitivity Reduction Technique for Discrete-Time Systems

The main drawback of using optimal control theory is the sensitivity to modeling errors. In this paper, a model error sensitivity reduction method is presented for discrete-time systems in an effort to solve the sensitivity to modeling error. The result is a decentralized control concept consisting of a collection of subsystem controllers, each independently charged with a subset of the system states. The resulting controllers are robust in the sense of having low sensitivity to errors in the parameters of the modeled system.