Real-time Fault Estimation for a Class of Discrete-Time Linear Parameter-Varying Systems

Abstract

In this article, we propose a diagnosis filter to estimate multiple fault signals for a class of discrete-time linear parameter-varying (LPV) systems. The design of such a filter can be formulated as an optimization problem to be solved recursively, while the system dynamics vary over time. The feasibility of such a problem is governed by given conditions on the detectability of the fault of interest. By formulating this convex problem in the quadratic programming framework, an analytical solution can be derived to approximate an optimal diagnosis filter at each time instance, which supports real-time implementation. The proposed methodology is illustrated and validated through an application in the lateral dynamics of an automated vehicle which is an often used example for this particular type of model class. The numerical results show that the diagnosis filter can decouple unknown disturbances and known or measurable parameter variations in the dynamics while detecting the unknown fault of interest.