Reinforced explorit on optimizing vehicle powertrains

Victor Parque, Masakazu Kobayashi, Masatake Higashi

研究成果: Conference contribution

8 被引用数 (Scopus)

抄録

How to build optimal vehicular powertrains? We study this question and propose an algorithm inspired by a domain-general design process. The basic idea is to interplay co-biasingly between the local approximations of discrete design and the global refinements of continuous parameters. The proposed method was evaluated to design powertrains of four types of vehicles: Series Hybrid Electric Vehicle(SHEV), Parallel Hybrid Electric Vehicle(PHEV), Fuel Cell(FC) and Electric Vehicle(EV). Simulation results show noticeable improvements on mileage per gas emissions over different study cases. To our knowledge, this is the first study aiming at designing vehicle powertrains considering the holistic point of view.

本文言語English
ホスト出版物のタイトルNeural Information Processing - 20th International Conference, ICONIP 2013, Proceedings
ページ579-586
ページ数8
PART 2
DOI
出版ステータスPublished - 2013 12 1
外部発表はい
イベント20th International Conference on Neural Information Processing, ICONIP 2013 - Daegu, Korea, Republic of
継続期間: 2013 11 32013 11 7

出版物シリーズ

名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
番号PART 2
8227 LNCS
ISSN(印刷版)0302-9743
ISSN(電子版)1611-3349

Other

Other20th International Conference on Neural Information Processing, ICONIP 2013
CountryKorea, Republic of
CityDaegu
Period13/11/313/11/7

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

フィンガープリント 「Reinforced explorit on optimizing vehicle powertrains」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル