Servo performance improvement through iterative tuning feedforward controller with disturbance compensator

Servo system is widely used in NC machines and its performance directly determines the precision of the machines. In most situations, the control structure for the servo system usually contains a cascaded P-PI feedback controller and a feedforward controller. This paper focuses on the feedforward controller parameters tuning to improve the servo performance. The feedforward controller consists of a model inversion and a parameterized disturbance model. Its parameters are tuned iteratively using the last cycle motion results. This method has the good extrapolation capability to the references and the performance improvement capacity. Moreover, it is easy to implement in real machines due to the simplicity and thus is of interest to control engineers. Experiments are carried out on an industrial prototype system. The results show that the proposed tuning method can improve the servo performance rapidly and the references are not required to keep the same during the tuning process.