The Fault Detection and Isolation Design for the Stratospheric Airship Based on the Interval Observer Method with Zonotope Analysis

This paper investigates a fault detection and isolation (FDI) scheme for the stratospheric airship (SA) flight control system under unknown external disturbances. In the design of the observer, the conventional nonlinear estimation error dynamic system is represented as Linear Parameter Varying (LPV) form mathematically. A Luenberger-type observer is designed to generate the residual for fault detection. Subsequently, the iterative zonotope method is employed to minimize the residual interval and the fault detection logical determination is developed. Furthermore, a bank of unknown input observers (UIO) is applied to realize the fault isolation issue with the zonotope analysis. Finally, the proposed method is applied on an SA control system and the result illustrates the effectiveness.