Reinforced explorit on optimizing vehicle powertrains

Victor Parque; Masakazu Kobayashi; Masatake Higashi

doi:10.1007/978-3-642-42042-9_72

Reinforced explorit on optimizing vehicle powertrains

Victor Parque, Masakazu Kobayashi, Masatake Higashi

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

10 Citations (Scopus)

Abstract

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.

Original language	English
Title of host publication	Neural Information Processing - 20th International Conference, ICONIP 2013, Proceedings
Pages	579-586
Number of pages	8
Edition	PART 2
DOIs	https://doi.org/10.1007/978-3-642-42042-9_72
Publication status	Published - 2013 Dec 1
Externally published	Yes
Event	20th International Conference on Neural Information Processing, ICONIP 2013 - Daegu, Korea, Republic of Duration: 2013 Nov 3 → 2013 Nov 7

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Number	PART 2
Volume	8227 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	20th International Conference on Neural Information Processing, ICONIP 2013
Country/Territory	Korea, Republic of
City	Daegu
Period	13/11/3 → 13/11/7

Keywords

Electric vehicle
Explorit
Fuel cell
Parallel hybrid electricvehicle
Reinforcement learning
Series hybrid electric vehicle

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-642-42042-9_72

Cite this

Parque, V., Kobayashi, M., & Higashi, M. (2013). Reinforced explorit on optimizing vehicle powertrains. In Neural Information Processing - 20th International Conference, ICONIP 2013, Proceedings (PART 2 ed., pp. 579-586). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8227 LNCS, No. PART 2). https://doi.org/10.1007/978-3-642-42042-9_72

Reinforced explorit on optimizing vehicle powertrains. / Parque, Victor; Kobayashi, Masakazu; Higashi, Masatake.
Neural Information Processing - 20th International Conference, ICONIP 2013, Proceedings. PART 2. ed. 2013. p. 579-586 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8227 LNCS, No. PART 2).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Parque, V, Kobayashi, M & Higashi, M 2013, Reinforced explorit on optimizing vehicle powertrains. in Neural Information Processing - 20th International Conference, ICONIP 2013, Proceedings. PART 2 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 2, vol. 8227 LNCS, pp. 579-586, 20th International Conference on Neural Information Processing, ICONIP 2013, Daegu, Korea, Republic of, 13/11/3. https://doi.org/10.1007/978-3-642-42042-9_72

Parque V, Kobayashi M, Higashi M. Reinforced explorit on optimizing vehicle powertrains. In Neural Information Processing - 20th International Conference, ICONIP 2013, Proceedings. PART 2 ed. 2013. p. 579-586. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2). doi: 10.1007/978-3-642-42042-9_72

Parque, Victor ; Kobayashi, Masakazu ; Higashi, Masatake. / Reinforced explorit on optimizing vehicle powertrains. Neural Information Processing - 20th International Conference, ICONIP 2013, Proceedings. PART 2. ed. 2013. pp. 579-586 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2).

@inproceedings{30a9bb11264348c981856e2de948171c,

title = "Reinforced explorit on optimizing vehicle powertrains",

abstract = "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.",

keywords = "Electric vehicle, Explorit, Fuel cell, Parallel hybrid electricvehicle, Reinforcement learning, Series hybrid electric vehicle",

author = "Victor Parque and Masakazu Kobayashi and Masatake Higashi",

year = "2013",

month = dec,

day = "1",

doi = "10.1007/978-3-642-42042-9_72",

language = "English",

isbn = "9783642420412",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

number = "PART 2",

pages = "579--586",

booktitle = "Neural Information Processing - 20th International Conference, ICONIP 2013, Proceedings",

edition = "PART 2",

note = "20th International Conference on Neural Information Processing, ICONIP 2013 ; Conference date: 03-11-2013 Through 07-11-2013",

}

TY - GEN

T1 - Reinforced explorit on optimizing vehicle powertrains

AU - Parque, Victor

AU - Kobayashi, Masakazu

AU - Higashi, Masatake

PY - 2013/12/1

Y1 - 2013/12/1

N2 - 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.

AB - 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.

KW - Electric vehicle

KW - Explorit

KW - Fuel cell

KW - Parallel hybrid electricvehicle

KW - Reinforcement learning

KW - Series hybrid electric vehicle

UR - http://www.scopus.com/inward/record.url?scp=84893388591&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84893388591&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-42042-9_72

DO - 10.1007/978-3-642-42042-9_72

M3 - Conference contribution

AN - SCOPUS:84893388591

SN - 9783642420412

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 579

EP - 586

BT - Neural Information Processing - 20th International Conference, ICONIP 2013, Proceedings

T2 - 20th International Conference on Neural Information Processing, ICONIP 2013

Y2 - 3 November 2013 through 7 November 2013

ER -

Reinforced explorit on optimizing vehicle powertrains

Abstract

Publication series

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this