Iterative learning control of Hamiltonian systems

This paper is concerned with iterative learning control of Hamiltonian systems, which is applicable to electromechanical systems. A novel iterative learning control scheme is proposed based the self-adjoint structure of the variational of those systems. This method does not require either the physical parameters of the target system nor the time derivatives of output signals. A concrete and effective learning algorithm for mechanical systems is also derived. Furthermore, experiments of a robot manipulator demonstrates the effectiveness of the proposed method.