Reinforcement learning in intelligent transportation systems: recent developments

Ye Yu; Quanquan Wang; Yukai Wang; Jiating Kuai

doi:10.1117/12.2613696

10 November 2021 Reinforcement learning in intelligent transportation systems: recent developments

Ye Yu, Quanquan Wang, Yukai Wang, Jiating Kuai

Proceedings Volume 12050, International Conference on Smart Transportation and City Engineering 2021; 120501J (2021) https://doi.org/10.1117/12.2613696
Event: 2021 International Conference on Smart Transportation and City Engineering, 2021, Chongqing, China

Abstract

With the increase in computing power and the development of reinforcement learning, the efficiency of the transportation system have been improving. Data-driven reinforcement learning algorithms have strong self-learning capabilities and adaptability to complex transportation systems and play a key role in the development of intelligent transportation systems. This paper reviews the basic concept of reinforcement learning firstly, including the terminology and methods which are divided into three categories: value-based, policy-based, and actor-critic methods. Secondly, the application and development of reinforcement learning in traffic signal control are introduced in the single-point control and the joint control of the multi-agent system. Then, this paper describes the development in autonomous driving separately from path planning and driving behavior. In the end, this paper concludes the challenges and problems faced by reinforcement learning methods in intelligent transportation systems.

Citation Download Citation

Ye Yu, Quanquan Wang, Yukai Wang, and Jiating Kuai "Reinforcement learning in intelligent transportation systems: recent developments", Proc. SPIE 12050, International Conference on Smart Transportation and City Engineering 2021, 120501J (10 November 2021); https://doi.org/10.1117/12.2613696

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available