Iterative robust control: Speeding up improvement through iterations

Iterative control has been widely studied in recent years as an efficient methodology for the design of highly-performing controllers of complex plants. The idea behind iterative design is that, when the plant is exceedingly complex, the design of the controller in one shot is hazardous. Instead, one can perform a sequence of closedloop identification and controller design steps, aiming at progressively learning how to increase the control performance through experience. In this paper, we introduce a new iterative control scheme which explicitly accounts for the presence of uncertainty in the plant description (iterative robust control). Our contention is that introducing robustness in iterative schemes permits one to quickly improve performance through steps, while preserving the robust stability of the closed-loop system.