A Decentralized Multi-agent Energy Management Strategy Based on a Look-Ahead Reinforcement Learning Approach

被引：8

作者：

Khalatbarisoltani A. ^{[1
]}

Kandidayeni M. ^{[2
,3
]}

Boulon L. ^{[1
]}

Hu X. ^{[4
,5
]}

机构：

[1] Hydrogen Research Institute, Department of Electrical and Computer Engineering, Université du Québec À Trois-Rivières, G8Z 4M3, QC

[2] Electric-Transport, Energy Storage and Conversion Lab (E-TESC), Université de Sherbrooke

[3] State Key Laboratory for Mechanical Transmission, Department of Automotive Engineering, Chongqing University, Chongqing

[4] Advanced Vehicle Engineering Centre, Cranfield University, Cranfield

来源：

SAE International Journal of Electrified Vehicles | 2021年 / 11卷 / 02期

关键词：

Energy management - Fuel cells - Reinforcement learning - Electric load flow - Powertrains - Energy efficiency;

D O I：

10.4271/14-11-02-0012

中图分类号：

学科分类号：

摘要：

An energy management strategy (EMS) has an essential role in ameliorating the efficiency and lifetime of the powertrain components in a hybrid fuel cell vehicle (HFCV). The EMS of intelligent HFCVs is equipped with advanced data-driven techniques to efficiently distribute the power flow among the power sources, which have heterogeneous energetic characteristics. Decentralized EMSs provide higher modularity (plug and play) and reliability compared to the centralized data-driven strategies. Modularity is the specification that promotes the discovery of new components in a powertrain system without the need for reconfiguration. Hence, this article puts forward a decentralized reinforcement learning (Dec-RL) framework for designing an EMS in a heavy-duty HFCV. The studied powertrain is composed of two parallel fuel cell systems (FCSs) and a battery pack. The contribution of the suggested multi-agent approach lies in the development of a fully decentralized learning strategy composed of several connected local modules. The performance of the proposed approach is investigated through several simulations and experimental tests. The results indicate the advantage of the established Dec-RL control scheme in convergence speed and optimization criteria. © 2021 SAE International Journal of Electrified Vehicles.

引用

共 63 条

[51]

Zhang K., Yang Z., Basar T., (2019)

[52]

Solano J., Jemei S., Boulon L., Silva L., Et al., Ieee Vts Motor Vehicles Challenge 2020 - Energy Management of a Fuel Cell/Ultracapacitor/Lead-Acid Battery Hybrid Electric Vehicle, pp. 1-6, (2019)

[53]

Martinez J.S., Hissel D., Pera M.-C., Amiet M., Practical Control Structure and Energy Management of a Testbed Hybrid Electric Vehicle, Ieee Transactions on Vehicular Technology, 60, pp. 4139-4152, (2011)

[54]

Kandidayeni M., Fernandez A.M., Boulon L., Kelouwani S., Efficiency Upgrade of Hybrid Fuel Cell Vehicles' Energy Management Strategies by Online Systemic Management of Fuel Cell, Ieee Transactions on Industrial Electronics, 68, pp. 4941-4953, (2020)

[55]

Watkins C.J., Dayan P., Q-Learning, Machine Learning, 8, pp. 279-292, (1992)

[56]

Chen H., Pei P., Song M., Lifetime Prediction and the Economic Lifetime of Proton Exchange Membrane Fuel Cells, Applied Energy, 142, pp. 154-163, (2015)

[57]

Herr N., Nicod J.-M., Varnier C., Jardin L., Et al., Decision Process to Manage Useful Life of Multi-Stacks Fuel Cell Systems under Service Constraint, Renewable Energy, 105, pp. 590-600, (2017)

[58]

Satyapal S., U.S. Department of Energy Hydrogen and Fuel Cell Technology Overview, (2018)

[59]

(2017)

[60]

Mongird K., Viswanathan V.V., Balducci P.J., Alam M.J.E., Et al., (2019)

← 1 2 3 4 5 6 7 →