Local observer-based multi-control and online validity estimation for multiple-model control of complex systems

The framework of the results presented here is the modelling approach with multiple models, for example the one used in a polytopic Takagi-Sugeno-like modelling and control of bounded non linear and uncertain systems where the base of models is composed of the vertices (or extreme) models. This paper proposes to sort out the problem of updating the state of each model encountered when estimating online the validities of each model in a multiple model approach. Moreover, the internal observations of each local-model state can be combined with these validities to control the system with a multi-control strategy, as the validity estimation is enhanced and more accurate. In the Takagi-Sugeno (TS) context, the online validities can be substituted to the contribution functions, giving better practical stability results.