Fault detection and identification for bimodal piecewise affine systems

This paper presents for the first time a fault detection and identification technique for bimodal piecewise affine (PWA) systems. A Luenberger-based observer structure is applied to the state estimation problem of the PWA system. The unknown value of the fault parameter is estimated by an observer equation obtained from a Lyapunov function. The design procedure is formulated as a set of linear matrix inequalities (LMIs) and guarantees global asymptotic stability of the estimation error, provided the norm of the input is upper and lower bounded by positive constants. The proposed method is applied to estimation of the amount of partial loss in control authority for a PWA model of a wheeled mobile robot (WMR).