Application of redundancy in observability and controllability towards fault tolerant system design

This work deals with fault tolerant control, in terms of a measure that aids in designing fault tolerant systems. The complexity of modern systems correlate with an increase in both performance needs and the risk of faults. This paper details a measure of redundancy that determines if it is possible for a system to recover from faulty actuators or sensors. This research creates a method to quantify the redundancy in a linear system to describe how many simultaneous faults the system can correct. It also identifies components of a system that are vulnerable to faults. Measures for both redundant observability, and controllability are presented. These measures do not guarantee fault identification is possible, but that the system is able to be fully reconstructed once fault isolation has been performed.