An optimal nonlinear predictive control based on a decoupled ARX-Laguerre multimodel

This paper deals with the representation of invariant, causal and stable systems on orthogonal Laguerre bases by combining the multimodel approach with the output and the input filtering principle. This representation allows the modeling of a large class of nonlinear systems and guarantees a significant parametric reduction. The resulting model of this representation is entitled ARX-Laguerre multimodel where the output is written as a linear combination of Laguerre filters and the Fourier coefficients. Thus, a new control technique of nonlinear systems, is proposed based on decoupled ARX-Laguerre multimodel.