Fractional order MPC design using improved state space model

Abstract

Abstract In order to improve the performance of control systems, a fractional order model predictive control (FMPC) algorithm using extended non-minimal state space (ENMSS) model is presented in this paper. For this propose, the fractional order model that describes the characteristics of practical processes more accurate than integer order model and the model predictive control (MPC) method based on ENMSS model that includes the state variable and output tracking error are considered. In this paper, the Grunwald-Letnikov (GL) definition is used to discretize the fractional order model and the fractional order cost function. This algorithm inherits the advantages of integer order extended non-minimal state space model predictive control (ENMSSMPC) with more adjustable parameters and better disturbance rejection, such that the controller tuning is more flexible. By comparing with the integer order ENMSSMPC, the results of the temperature control of an industrial heating furnace indicate the performance improvement of the proposed method.