Some remarks on optimal output regulation for weakly dual redundant plants