Two effective control algorithms for a pneumatic system

Performing accurate control for pneumatic systems is very difficult mainly due to the compressibility of airflow, friction forces and actuator saturation. This paper implements the anti-windup control and Generalized Predictive Control (GPC) for a pneumatic system based on its third order ARX model from neural network trained with Levenberg-Marquard method. Although the anti-windup controller with Ragazzini method overcomes the actuator saturation effectively, the time-delay still exists to some extent. To resolve the time-delay problem, GPC algorithm is applied for the control of the pneumatic system. The performance of the designed GPC controller is very impressive for its fast and good tracking with high accuracy.