Extended Kalman filter based unconstrained model predictive control of a complex nonlinear system: the quadruple tank process

Abstract

This paper proposes the model predictive controller (MPC) based on the Kalman filter for a complicated nonlinear system—the quadruple tank process (QTP). The control of a multivariable and nonlinear system like a QTP is a difficult job. A number of nonlinear design techniques are implemented to ameliorate the pursuit performance of the QTP, however, the nonlinear techniques make implementation composite and computationally unsuitable. In this work, an unconstrained MPC is planed for the QTP experiences and it is controlled for both minimum and non-minimum sentence configurations in order to follow the wanted track. Its performance can be damaged once system is pass from minimum to non-minimum phase region and inversely. The unknown states required for model predictive control design are rebuilt using an extended Kalman filter. The design of model predictive control and extended Kalman filter is based on the QTP and the achievement of the proposed controller is checked for the monitoring of references. All results of simulation are affected using the MATLAB software. The results of the simulation show the capability and power of the suggested controller in respect of monitoring the trajectory and state estimation.