Optimal energy management for a plug-in hybrid electric vehicle: Real-time controller

• Published in: Proceedings of the 2010 American Control Conference pp. 5037 - 5042
• Main Authors: Xiao Lin, Banvait, H, Anwar, S, Yaobin Chen
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.06.2010
• Subjects: Analytical models; Batteries; Energy management; Hybrid electric vehicles; Mechanical power transmission; Neural networks; Optimal control; Power system management; Power system modeling; Real time systems
• ISBN: 9781424474264; 1424474264
• ISSN: 0743-1619; 2378-5861
• DOI: 10.1109/ACC.2010.5530731

A converted Toyota Prius 2007 plug-in hybrid electric vehicle (PHEV) uses additional large capacity battery, so that it can enhance the pure electric drivability and increase its electric range. To accomplish real time energy distribution management system in this PHEV, firstly, a specific model is established which contains most of the powertrain properties and partly vehicle dynamics. Secondly, an optimal control problem with inequality constraints is analyzed and formulated mathematically. Thirdly, the particle swarm optimization (PSO) is applied to search for global near-optimum at each time interval. However, PSO is time-consuming so it can be used only as an off-line controller. To overcome this drawback neural network is designed to get sub-optimal real-time controller by employing the near-optimal results obtained from aforementioned PSO. Finally, all the results which are obtained from the PSO and the neural network are then simulated on PSAT (Powertrain System Analysis Toolkit), a standard vehicle simulation software. In the end, the results are compared which show that the real-time controller using neural network could get good sub-optimal results without sacrificing any vehicle performance of vehicle.