Adaptive control of sliding table position using feedback linearization neural networks

Abstract

Control of sliding table position in mechanical work machines is very important, especially In CNC machines because it requires high accuracy. In most mechanical machines, speed or position controllers have used PID controllers, so they are less adaptable to the noise as well as the parameter changes of the model. This paper proposes a solution that uses a feedback linearization control and a nonlinear autoregressive-moving (NARMA-L2) adaptive neuron controller. The idea of this type of controller is to approximate the nonlinear dynamical system into a linear dynamical system. Initially, the system model is built, then the NARMA-L2 controller is used to identify the system and generate control signals for the object. The results of this study show that the ability to adapt to noise and the change in model parameters during operation of the NARMA-L2 is better than that of the PID controller.