Fault diagnosis of hyperbolic distributed parameter systems

We consider parameter estimation techniques for detecting changes of a distributed nature due to faults, in a plant whose dynamics are governed by a wave equation. We examine changes in the distributed stiffness term, which provides a very simple and practical failure model for distributed parameter systems. Using Lyapunov redesign methods, extended and applied to hyperbolic distributed parameter systems, a stable learning scheme for fault diagnosis is developed. The resulting non-destructive fault diagnosis scheme is able not only to detect the occurrence of a fault but also to identify the location and shape of the distributed nature of the fault. A numerical algorithm is provided for the implementation of the detection scheme and simulation studies are used to illustrate the applicability of the theoretical results.