Multivariable Controller Design for a Hydraulic Screw-down System with Experiments

In this paper the synchronization issue of two hydraulic cylinders is considered in hydraulic screw-down system. Accounting for the two cylinders system with complicated dynamics characterized by nonlinear, time-varying and load coupling properties, a decoupling arithmetic based on dynamic ANN inverse system is proposed. Firstly, the reversibility of the system is analyzed and the inverse dynamic is constructed based on a static neural network and an integrator. Secondly, the decoupled pseudo-linear system can be obtained through combining the inverse dynamic model with the original system in series. Aim to pseudo-linear system NN-PID controller is proposed to satisfy high performance control. Experiments were conducted to verify the performance of the proposed controller and have shown that the system has good transient response on 300-reversible cold mill.