Fictitious reference iterative tuning of cascade controllers for non-minimum phase systems

Abstract

In this paper, we address a new parameter tuning method of cascade controllers systems for linear time-invariant systems with unstable zeros, i.e., non-minimum phase systems. Under the situation where an appropriate model of a plant is unknown and only measured data are available, the proposed method in this paper yields the optimal parameter that achieves the desired tracking property of the cascade control systems. Here, we apply fictitious reference iterative tuning (FRIT). This method is a tuning method that yields the desired controller parameter with only one-shot experiment data. Particularly, noting that a system is non-minimum phase and unknown, we parameterize the non-minimum phase part to the desired transfer function because the non-minimum phase property is related to the limit of the desired tracking property.