Experimental validation of model predictive control for nonlinear system with delay [articol]

Abstract

Industrial processes have inherent nonlinearity, process delay, transportation delay, and might suffer from inverse response. Every control action is performed based on the error calculated at that instant. But in systems encountering transportation delay, this is not possible because, at every instant only the past information could be made available to the controller for taking corrective measures. Predictive controllers could well handle this issue. They predict the system behaviour by using the past behaviours of the plant and so they could overcome the issues faced by transportation delay. This paper presents the real-time implementation of model predictive controller for a level process with delay. The performance of this controller for this system when transportation delay is intentionally included to the system is compared with the performance of the classical PID controller.