Robust fault estimation in non-infinitely observable systems using two sliding mode observers

What is it about?

Existing methods to estimate faults in descriptor systems that are not infinitely observable require the system to satisfy stringent mathematical conditions. By utilising two sliding mode observers in cascade (as opposed to a single observer), the conditions can be relaxed, thus increasing the applicability of the scheme. Furthermore, the sliding mode observers are designed in a way such that the fault estimate will be least corrupted by disturbances affecting the system.

Featured Image

Photo by Caspar Camille Rubin on Unsplash

Photo by Caspar Camille Rubin on Unsplash

Why is it important?

Current methods for fault estimation in descriptor systems that are not infinitely observable place strict requirements on the system. These requirements could potentially translate to a larger number of sensors being required, thus increasing the complexity and cost of the fault estimation scheme. Our findings relax these requirements, lowering the number of sensors required. The presented scheme would therefore be simpler and comparatively cost less. Furthermore, by designing the observer scheme to be robust against disturbances, the number of fault channels can be reduced since we can model them as disturbances. This would again contribute to the reduction of the number of sensors required.

Perspectives