Intelligent position control for pneumatic servo system based on predictive fuzzy control

In this research, a novel method of predictive fuzzy control combined with a neural network is proposed for a pneumatic servo system. With the proposed method, the future outputs of the plant are predicted by using an imaginary plant in the predictive fuzzy scheme. The error between the desired value and the output of the imaginary plant, and the change in this error is applied as the input signals of the predictive fuzzy controller to generate a direct control input. In the proposed method, the neural network is introduced to construct the imaginary plant model of the target pneumatic servo system. To obtain teaching signals for the neural network scheme, a pre-test based on Proportional-integralderivative (PID) control is implemented, and the results are applied as the teaching signals. With the imaginary plant model, the effectiveness of the proposed method using predictive fuzzy control is experimentally verified, and satisfactory position control using the proposed method is confirmed. (C) 2019 Published by Elsevier Ltd.