Performance Evaluation of Model Predictive Control for Three Interacting Tank Process Based on its Lower Order Integer and Non-Integer Models

Abstract

This paper establishes the superiority of fractional order model reduction of a higher order system over integer order in the performance evaluation of a very popular model based control technique nowadays i.e. Model Predictive Control (MPC). The procedure of model order reduction is adopted to reduce the complexity in controller design. A MPC strategy is developed using MATLAB MPC tool for a benchmark level control problem of a three interacting tank process, where the level of third tank is controlled by manipulating input flow rate to the first tank. In MATLAB, various fractional order systems are defined with the help of FOMCON toolbox. Different simulation results show enhancement in the performance of MPC based on non-integer order model as compared integer order model.