Reliable and Efficient Lane Changing Behaviour for Connected Autonomous Vehicle through Deep Reinforcement Learning

The establishment of future intelligent transport systems is dependable on the reliable and seamless function of Connected and Autonomous Vehicles (CAV). Reinforcement learning (RL), which allows autonomous vehicles (AVs) to learn an ideal driving strategy through constant contact with the environment, plays a significant part in the decision-making process of autonomous driving (AD). The networking of CAV is advantageous since it allows for the transmission of traffic-related data to vehicles via Vehicle-to-External (V2X) communication. Recognition and anticipation of driving behaviour are critical for avoiding collisions because they can provide useful information to other drivers and vehicles. The fundamental challenge in developing CAV is the construction of an autonomous controller that can effectively perform close real-time control selections, such as a fast acceleration while merging onto a highway and rapid speed adjustments in stop-and-go traffic congestion. CAV driving behaviours can be considerably improved by utilizing shared information, resulting in more accountable, intelligent, and efficient driving. In the present work, a deep reinforcement learning approach is proposed that integrates the information gathered through connectivity capabilities and sensing from neighbour automobiles in the vicinity of CAV. The fused information is used for providing safe and cooperative lane-changing behaviour. The deployment of an algorithm in CAV is expected to improve the transportation safety of CAV driving behaviours.

• Record URL:
  https://doi.org/10.1016/j.procs.2023.01.090
• Record URL:
  http://www.sciencedirect.com/science/article/pii/S187705092300090X
• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/issn/18770509
• Supplemental Notes:
  • © 2023 The Author(s). Published by Elsevier B.V. Abstract reprinted with permission of Elsevier.
• Authors:
  • Alagumuthukrishnan, S
  • Deepajothi, S
  • Rajasekar, Vani
  • Velliangiri, S
• Conference:
  • International Conference on Machine Learning and Data Engineering (ICMLDE 2022)
  • Location: Dehradun , India
  • Date: 2022-9-7 to 2022-9-8
• Publication Date: 2023

Language

• English

Media Info

• Media Type: Web
• Features: Figures; References; Tables;
• Pagination: pp 1112-1121
• Serial:
  • Procedia Computer Science
  • Volume: 218
  • Issue Number: 0
  • Publisher: Elsevier
  • ISSN: 1877-0509
  • Serial URL: http://www.sciencedirect.com/science/journal/18770509
• Publication flags:

  Open Access (libre)

Subject/Index Terms

• TRT Terms: Algorithms; Autonomous vehicles; Connected vehicles; Lane changing; Machine learning; Vehicle to infrastructure communications
• Subject Areas: Highways; Vehicles and Equipment;

Filing Info

• Accession Number: 01875512
• Record Type: Publication
• Files: TRIS
• Created Date: Mar 13 2023 12:53PM