Optimal trajectory tracking control of servo hydraulic die cushion based on adaptive iterative learning control

Servo hydraulic die cushions are widely used in high-end press systems to control the blank-holder force and movement of the workpieces. The motion control performance of the servo hydraulic die cushion is a key factor for improving the production efficiency. As a kind of electro-hydraulic servo systems, servo hydraulic die cushions have many existing strategies for motion control. However, many control parameters need to be tuned under varied working conditions for most control methods to satisfy the performance requirements. In this study, a novel adaptive iterative learning controller based on backstepping method is proposed for motion control of servo hydraulic die cushions. The controller contains iterative learning laws, a parameter adaptation law and a robust dynamical control term to handle all kinds of uncertainties. The experiments conducted under typical working conditions show that the proposed controller can work with one set of parameters and the performance requirements can be satisfied after several learning iterations. Furthermore, the motion trajectory is planned by a Legendre-Gauss-Radau quadrature orthogonal collocation method to satisfy the terminal state constraints and inequality path constraints.