A deep reinforcement learning approach for integrated optimization of train scheduling and rolling stock circulation planning

Xiaoli Zhao

doi:10.1016/j.cie.2025.111784

A deep reinforcement learning approach for integrated optimization of train scheduling and rolling stock circulation planning

2025-12-25

0

PRE

AI

赵

赵晓丽 (Xiaoli Zhao)

D

Dewei Li *

X

Xinyu Bao

DOI：10.1016/j.cie.2025.111784

Original

Original request for help

Save

Abstract

En 中文

• Proposes a deep reinforcement learning framework for metro train timetable and rolling stock circulation planning. • Develops a hybrid action space environment to coordinate discrete and continuous operational decisions. • Designs a potential-based reward shaping mechanism to enhance learning efficiency. • Handles operational constraints via action masking and action embedding mechanisms.

A deep reinforcement learning approach for integrated optimization of train scheduling and rolling stock circulation planning

Abstract

Journal

Organization

Cited Papers

Citing Papers