Evpool: New Power Management Framework for Data Centers

Author(s):  
Fengqian Gao ◽  
Qiang Huang ◽  
Rui Wang
Author(s):  
Jie Liu ◽  
Feng Zhao ◽  
Xue Liu ◽  
Wenbo He

Author(s):  
Thant Zin Oo ◽  
Nguyen H. Tran ◽  
Shaolei Ren ◽  
Choong Seon Hong

Author(s):  
Qishen Zhao ◽  
Tianheng Feng ◽  
Dongmei Chen ◽  
Wei Li

Abstract Electrification of locomotive with hybridized fuel-cell, battery and supercapacitor has drawn much attention from both the academia and industry. Unlike traditional powertrain, hybrid powertrain consists of multiple power sources with a complex drivetrain structure, various efficiency performance, and different dynamics. Therefore, it is necessary to develop a power management strategy to make sure each power source operates under a quasi-optimal condition and maximize the overall powertrain efficiency. This paper presents the development of a power management framework for a novel hybrid locomotive consisting of PEM fuel cell, battery, and supercapacitor. Both the equivalent consumption management strategy (ECMS) and the stochastic dynamic programming (SDP) are applied to solve for the optimal power split strategy. The resulted power management strategy is presented in the form of policy maps, which makes it convenient for real-time in-vehicle implementations. Simulation results indicate that the SDP demonstrates advantages over the ECMS in terms of equivalent hydrogen consumption over typical locomotive driving cycles.


2019 ◽  
Vol 2019 (S2) ◽  
pp. S1-S26
Author(s):  
Kanad Ghose ◽  
Dale Becker

Abstract Invited Session on HETEROGENEOUS INTEGRATION ROADMAP - Heterogeneous Integration for HPC and Data Centers. This TWG focuses on the system-level implications related to performance, power management, security, power distribution issues and others


Computing ◽  
2020 ◽  
Vol 102 (7) ◽  
pp. 1717-1741
Author(s):  
Qiang Wang ◽  
Haoran Cai ◽  
Qiang Cao ◽  
Fang Wang

Sign in / Sign up

Export Citation Format

Share Document