Algebraic H $ınfty$ design of Higher-Order Iterative Learning Controllers