A new design method for reliable fuzzy control

This paper concerns about the reliable control design problem for the fuzzy hyperbolic model (FHM) with time-varying delay and actuator faults. FHM is a universal approximate as well as a global model and can be easily derived from a set of fuzzy rules. It can also be seen as a feedforward neural network model, and so the model parameters can be identified by BP-algorithm. The method for designing a reliable hyperbolic controller is presented in terms of linear matrix inequalities by applying a descriptor model transformation of the system and free-weighting matrices. The resulting closed-loop systems are reliable in the sense that asymptotic stability is achieved when the system is in the normal and in the presence of actuator fault cases.