A Comprehensive Survey on Multi-Agent Reinforcement Learning for Connected and Automated Vehicles

Area of Science:

• Intelligent Transportation Systems
• Artificial Intelligence
• Robotics

Background:

• Connected and Automated Vehicles (CAVs) necessitate complex, simultaneous task management, including motion planning and traffic control.
• Multi-Agent Reinforcement Learning (MARL) offers a promising framework for addressing these intricate control problems.
• Existing research lacks a comprehensive overview of MARL applications in CAVs, hindering further development.

Purpose of the Study:

• To provide a comprehensive survey of Multi-Agent Reinforcement Learning (MARL) applications in Connected and Automated Vehicles (CAVs).
• To analyze current research trends, identify key challenges, and propose future research directions in MARL for CAVs.

Main Methods:

• A classification-based analysis of existing research papers on MARL for CAVs.
• Identification and categorization of current developments and research directions.
• Discussion of challenges and potential future research avenues.

Main Results:

• The survey categorizes current MARL research for CAVs, highlighting diverse approaches and applications.
• Key challenges in implementing MARL for complex CAV tasks are identified.
• Potential future research areas are proposed to address identified challenges.

Conclusions:

• MARL is a critical technology for enabling sophisticated functionalities in CAVs.
• Further research is needed to overcome current challenges and fully realize MARL's potential in intelligent transportation systems.
• This survey serves as a valuable resource for researchers and practitioners in the field of MARL for CAVs.