Modelling and Controlling the Level ofNonlinear Process via Diverse ControlStrategies

[摘要] In process industries, the level of the liquid must be controlled and it is a mandatory process. But controlling a non-linear process is a highly complicated and complex process. Many process industries employs conical tank because of its non-linear shape, because of its constantly changing cross section, with respect to height controlling. For each operating region, the linear portion of the nonlinear process is determined as first order plus dead time model (FOPD). Different controller scheme such as conventional PID controllers based on Ziegler-Nicholas (ZN-PID) method, Chien, Hrones and Reswich (CHR-PID) method, Tyres-Luyben (T-L-PID) method, Internal Model Control (IMC-PID) and Model Predictive Controller (MPC) method were proposed and their responses was compared. The main objective of the controller is to provide a minimum settling time, peak overshoot and rise time and to provide a minimum performance error. Among the various controllers, Model Predictive Control (MPC) satisfies the above criteria.

[关键词] Non-Linear process;MPC;IMC;Conical tank