Gan Chen

μ-analysis and synthesis of state feedback systems based on multipliers and LMI's

This paper provides a μ analysis and synthesis method of state feedback systems. First, we derive a sufficient condition for the dynamical system to have μ less than a specified value γ by using parameter-dependent multipliers and the positive real lemma. It requires no higher order multiplier and no frequency sweep while being less conservative than former methods. Second, based on this result, we give an LMI condition for the existence of a state feedback gain which achieves that μ of the closed loop system is less than given γ. Finally, an illustrative numerical example is given.

Mixed-/spl mu/ analysis for dynamical systems using descriptor form

This paper provides a new mixed-/spl mu/ analysis method for dynamical systems without the frequency sweep. First, we derive a necessary and sufficient condition for the closed loop system with dynamical system and real uncertainties to have /spl mu/ less than a specified value /spl gamma/ by using descriptor form. Second, applying the sector transformation and the strictly positive real lemma to this result, we give a computable LMI condition in the state space for the mixed-/spl mu/ analysis problem. The proposed method can treat all uncertainties as time invariant, and it may be less conservative than former ones including the D-G scaling. Finally, an illustrative numerical example is given.

Identification of continuous-time systems via genetic algorithm

In this paper, we propose an identification method for continuous-time systems by the genetic algorithm. It is well known that the prediction error method (PEM) is conventionally used as an identification method, however, it is applicable only to discrete-time models. By introducing the genetic algorithm, the proposed method is applicable to continuous-time models. Through identifying a continuous-time model, we discuss problems in the PEM and availability of the proposed method.