Said H. Esfahani - School of Electrical Engineering, Shahr-e-Kord University, Shahr-e-Kord, Iran.

This paper is concerned with the problem of suboptimal outputfeedback steady-state L -gain control design for nonlinear systemsunder the effects of persistent bounded disturbances. The initialnonlinear system is approximated by a suitable T-S fuzzy model. Itis assumed that the information of premise variables in the T-Sfuzzy model is not accessible through the measurement equations.In contrast to the existing results in the literature, the controller inthis paper does not have an observer-based structure. Moreover, itis not in the form of a Parallel Distributed Compensation (PDC)control laws. In this paper, the T-S fuzzy model is first transformedto an equivalent linear uncertain system. The control problem isthen formulated in the form of a single robust steady-state L -gaincontrol law for the uncertain system. The single controller can beeasily found by a search over the solutions of two Linear MatrixInequalities (LMIs). However, four positive scalars should bedetermined in advance. Hence, a particular numerical method isdeveloped to find the feasible solutions. Moreover, the numericalmethod can be easily extended to find the suboptimal control law.One illustrative example shows the effectiveness of the obtainedresults.

L -gain control, T-S fuzzy systems, Persistent boundeddisturbances, Immeasurable premise variables, Robust control