Data-Driven Reinforcement Learning–Based Real-Time Energy Management System for Plug-In Hybrid Electric Vehicles

Transportation Research Record - Tập 2572 Số 1 - Trang 1-8 - 2016
Xuewei Qi1, Guoyuan Wu2, Kanok Boriboonsomsin2, Matthew Barth1, Jeffrey Gonder3
1Department of Electrical and Computer Engineering, University of California, Riverside, 1084 Columbia Avenue, Riverside, CA 92507
2CE-CERT, University of California, Riverside, 1084 Columbia Avenue, Riverside, CA 92507
3National Renewable Energy Laboratory, 15013 Denver West Parkway, Golden, CO 80401

Tóm tắt

Plug-in hybrid electric vehicles (PHEVs) show great promise in reducing transportation-related fossil fuel consumption and greenhouse gas emissions. Designing an efficient energy management system (EMS) for PHEVs to achieve better fuel economy has been an active research topic for decades. Most of the advanced systems rely either on a priori knowledge of future driving conditions to achieve the optimal but not real-time solution (e.g., using a dynamic programming strategy) or on only current driving situations to achieve a real-time but nonoptimal solution (e.g., rule-based strategy). This paper proposes a reinforcement learning–based real-time EMS for PHEVs to address the trade-off between real-time performance and optimal energy savings. The proposed model can optimize the power-split control in real time while learning the optimal decisions from historical driving cycles. A case study on a real-world commute trip shows that about a 12% fuel saving can be achieved without considering charging opportunities; further, an 8% fuel saving can be achieved when charging opportunities are considered, compared with the standard binary mode control strategy.

Từ khóa


Tài liệu tham khảo

Bureau of Transportation Statistics (BTS). http://www.bts.gov/publications/national_transportation_statistic.

2015, DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990–2013

10.1109/TITS.2013.2294342

10.1016/j.egypro.2014.01.100

10.1049/iet-its.2014.0075

10.3390/en6115656

10.1109/ChiCC.2014.6895690

10.1109/ACC.2013.6580774

10.1109/TCST.2013.2278684

10.3182/20080706-5-KR-1001.00785

10.1109/TVT.2014.2336378

10.1109/TITS.2013.2294723

Qi X., 2014 IEEE 17th International Conference on Intelligent Transportation Systems, 2480

O’Keefe M. P., 2006, Dynamic Programming Applied to Investigate Energy Management Strategies for a Plug-In HEV

10.1016/j.jpowsour.2013.09.085

Lin X., 2010, American Control Conference, 5037

10.1016/j.jfranklin.2014.07.009

10.1109/IVS.2015.7225722

10.1109/TCST.2008.919447

Bellman R. E., Dynamic Programming

10.1002/9781118029176

10.1016/j.trc.2009.04.005

Sutton R. S., 1998, Reinforcement Learning: An Introduction

California Performance Measurement System. http://pems.dot.ca.gov/. Accessed July 7, 2015.

Thông tin
Thông tin xuất bản

Transportation Research Record

Tập 2572 Số 1

1-8

Thông tin tác giả