Dual Iterative Learning Control for Multiple-Input Multiple-Output Dynamics with Validation in Robotic Systems