Real-time and optimal energy management strategy via explicit model predictive control for small fuel cell hybrid vehicles

Abstract

An energy management strategy (EMS) is the crucial factor in achieving vehicle driving economy, energy source durability and real-time energy allocating. This paper proposes the real-time and optimal EMS for hydrogen fuel cell/power battery-based vehicles with low cost and limited computing power via explicit model predictive control (EMPC). Based on the establishment of prediction model, cost function and constraints, the linear piecewise affine law of EMPC is obtained by offline computation utilizing multi-parameter quadratic programming theory and Matlab's Multi-parameter Toolbox 3. The co-simulation results with Matlab and Advisor demonstrate the strategy reduces equivalent hydrogen consumption by 18.93 %, 2.87 %, 0.76 % and 11.79 %, 5.51 %, 1.22 % respectively compared to the power-following strategy, the power-following fuzzy strategy, and the conventional model predictive control strategy under two standard cycle conditions; and its average computing time within a 1-s sampling period is only 6.48 ms, while those of the other strategies are 1.33, 1.83, and 10.44 times, respectively. Therefore, the strategy provides notable fuel economy and excellent real-time control performance for this type of fuel cell vehicle.