Multi-mode modelling with linear dynamic models using set-membership tools
Résumé
The aim of this work is to model and to simulate a SISO switching system with several operating modes by a set of linear models. Each operating mode is modelled individually by several parallel dynamic models with varying moving horizons in which model parameters are bounded variables. In each operating mode a static model is also added in parallel to dynamic ones to overcome memory effect and they all together should explain system behaviors. This model can be used in different domains such as process safety analysis and fault detection in control engineering where one needs to predict all allowed and normal system behaviors.