Modelling and monitoring of changes in dynamical systems
Albert Benveniste, Michèle Basseville, George V. Moustakides · 1986
We present a systematic approach for the design of change detection and model validation algorithms for dynamical systems. We show how to associate to any identification algorithm a change detection and a model validation procedure, which are optimal in some asymptotic meaning. The foundations of our method go back to the «asymptotic local» approach in statistics, and our method generalizes this approach.