Parameter complexity reduction of multivariable systems using genetic algorithm

In this paper, we propose to present the modelling of multivariable Autoregressive with exogenous input (ARX) model by using Laguerre functions. The optimal approximation of the resulting model, entitled MIMO ARX Laguerre model, is ensured when the poles characterizing each Laguerre bases are optimal. An evolutionary method is proposed for the optimization of the Laguerre poles from input/output measurements. The effectiveness of this method is tested on numerical example.