Zur Hauptnavigation wechseln Zur Suche wechseln Zum Hauptinhalt wechseln

Multi-agent Soft Actor-Critic with Coordinated Loss for Autonomous Mobility-on-Demand Fleet Control

  • Technische Universität München

Publikation: Beitrag in Buch/Bericht/KonferenzbandKonferenzbeitragBegutachtung

Abstract

We study a sequential decision-making problem for a profit-maximizing operator of an autonomous mobility-on-demand system. Optimizing a central operator’s vehicle-to-request dispatching policy requires efficient and effective fleet control strategies. To this end, we employ a multi-agent Soft Actor-Critic algorithm combined with weighted bipartite matching. We propose a novel vehicle-based algorithm architecture and adapt the critic’s loss function to appropriately consider coordinated actions. Furthermore, we extend our algorithm to incorporate rebalancing capabilities. Through numerical experiments, we show that our approach outperforms state-of-the-art benchmarks by up to 12.9% for dispatching and up to 38.9% with integrated rebalancing.

OriginalspracheEnglisch
TitelLearning and Intelligent Optimization - 19th International Conference, LION 19 2025, Proceedings
Redakteure/-innenYingqian Zhang, Milan Hladik, Hossein Moosaei
Herausgeber (Verlag)Springer Science and Business Media Deutschland GmbH
Seiten1-15
Seitenumfang15
ISBN (Print)9783032091550
DOIs
PublikationsstatusVeröffentlicht - 2026
Veranstaltung19th International Conference on Learning and Intelligent Optimization, LION 2025 - Prague, Tschechische Republik
Dauer: 15 Juni 202519 Juni 2025

Publikationsreihe

NameLecture Notes in Computer Science
Band15744 LNCS
ISSN (Print)0302-9743
ISSN (elektronisch)1611-3349

Konferenz

Konferenz19th International Conference on Learning and Intelligent Optimization, LION 2025
Land/GebietTschechische Republik
OrtPrague
Zeitraum15/06/2519/06/25

Fingerprint

Untersuchen Sie die Forschungsthemen von „Multi-agent Soft Actor-Critic with Coordinated Loss for Autonomous Mobility-on-Demand Fleet Control“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren