Unit Prediction Horizon Constrained Model Predictive Control for Plug-in Hybrid Electric Vehicle

Abstract

Plug-in hybrid electric vehicles (PHEVs) provide a good alternative in achieving better performance and in the reduction of harmful gas emissions. The hybrid energy storage system (HESS) and the integrated charging unit constitute the PHEV under consider ation. To meet the load demands, the proposed HESS is a coupled system comprising a fuel cell, high energy density battery, and high-power density super-capacitor. On board charging involves the utilization of a DC–DC buck converter and an uncontrolled rectifier and two buck-boost converters are utilized to facilitate a smooth transfer of energy. A rule-based supervisory controller has been implemented, which takes into account the state of charge of energy sources and also the total power inflow of the power sources. A model predictive control (MPC) technique is implemented to ensure that PHEVs function smoothly in terms of regulation of DC bus voltage and tracking of current. Unit prediction horizon i.e. only one step ahead looking into the future is considered for the MPC to make it computationally less expensive. MPC is designed in such a way that its performance is close to our favorite linear controller which in our case is the H∞ controller. The inverse optimal control technique is used for determining the weight matrices for MPC. The proposed controller has been simulated using MATLAB and performance is compared with linear matrix inequality based H∞ optimal control to prove that its performance is similar to that of H∞ controller. Also, the performance comparison is made between the designed MPC and the non-linear control approach i.e. sliding mode control to show that it achieves comparable or superior performance to the non-linear controller. The real-time performance of H∞ based MPC is verified using controller hardware in the loop experimental setup.