Nonlinear Robust Identification Using Multiobjective Evolutionary Algorithms

Abstract

In this article, a procedure to estimate a nonlinear models set (Θ _p ) in a robust identification context, is presented. The estimated models are Pareto optimal when several identification error norms are considered simultaneously. A new multiobjective evolutionary algorithm $\epsilon\nearrow - MOEA$ has been designed to converge towards Θ$_{P}^{\rm \star}$, a reduced but well distributed representation of Θ _P since the algorithm achieves good convergence and distribution of the Pareto front J(Θ). Finally, an experimental application of the $\epsilon\nearrow - MOEA$ algorithm to the nonlinear robust identification of a scale furnace is presented. The model has three unknown parameters and ℓ_∞ and ℓ₁ norms are been taken into account.