Computational burden reduction in set-membership identification of Wiener models