Model-based Monitoring using Dynamic Uncertainty Space Partitioning

B. Rinner and U. Weiss (Austria)


Fault Detection; Identification; Imprecise Models; Self calibrating Monitoring


Monitoring gains importance for many technical systems such as robots, production lines or anti lock brakes. A monitoring system for technical systems must be able to deal with incomplete knowledge of the supervised system, to process noisy observations and to react within predefined time windows. This paper presents a new approach to monitoring technical systems based on imprecise models. Our ap proach repeatedly partitions the uncertainty space of an im precise model and checks the derived model’s state for con sistency with the measurements. Inconsistent partitions are then refuted resulting in a smaller uncertainty space and a faster fault detection. We demonstrate our monitoring ap proach on a non-trivial heating system where we achieve a significant reduction of the fault detection time compared to conventional interval methods.

Important Links:

Go Back