Home > Published Issues > 2018 > Volume 7, No. 3, May 2018 >

Comparative Analysis of MPC Based on Integer and Non-integer Order Models: Case Studies

Abdul Wahid Nasir, Idamakanti Kasireddy, and Arun Kumar Singh
Department of Electrical & Electronics Engineering, National Institute of Technology (NIT) Jamshedpur, India

Abstract— The present work makes use of the fractional order modeling in realizing Model Predictive Control (MPC) for some processes i.e. DC motor and Water Bath system, whose models are derived from their open loop experimental data. The process behavior is investigated from experimental setup, then for the controller performance evaluation, simulation based analysis is made. Firstly, open loop experimental data is collected by applying a step change in the manipulating variable. Based on these data, integer and non-integer order model parameters of the processes are estimated using Genetic Algorithm (GA) respectively, by minimizing Integral of Squared Error (ISE) between open loop step data and model response. Once the models are available, a model based control technique MPC is designed and simulated using MATLAB MPC toolbox. Before using fractional order model for MPC design, it is firstly converted to equivalent higher integer order model using Oustaloup’s recursive approximation. Since the performance of such type of control technique depends on the accuracy of model of the plant. Therefore, MPC based on non-integer model give better performance as compared to integer order model, as former is able to capture the model dynamics more accurately than later. Different simulations performed in MATLAB also approves the same. 

Index Terms—Fractional order modeling, model predictive control, water bath system, DC motor, model identification, genetic algorithm.

Cite: Abdul Wahid Nasir, Idamakanti Kasireddy, and Arun Kumar Singh, "Comparative Analysis of MPC Based on Integer and Non-integer Order Models: Case Studies," International Journal of Mechanical Engineering and Robotics Research, Vol. 7, No. 3, pp. 264-272, May 2018. DOI: 10.18178/ijmerr.7.3.264-272