Gain-scheduled control synthesis by using filtered scheduling parameters

This paper is concerned with design of gain-scheduled (GS) controllers that depend on filtered scheduling parameters. One of the most sophisticated design method, which is based on parameter-dependent Lyapunov matrices, ends up with GS controllers involving differentiation of the scheduling parameter. On the other hand, proposed GS controllers with filtered parameters avoid derivative terms of the scheduling parameter. We show that our synthesis method has guaranteed performance level less than that of GS controllers designed via constant Lyapunov matrix. Moreover, if the filter has time-constant short enough, the performance level via the filtered scheduling parameter is close to that of conventional GS controller that needs differentiation of the parameter.