A Complete Framework for a Behavioral Planner with Automated Vehicles: A Car-Sharing Fleet Relocation Approach.

Area of Science:

• Robotics and Autonomous Systems
• Urban Mobility Solutions
• Intelligent Transportation Systems

Background:

• Automated vehicle research primarily focuses on highways, leaving complex urban environments challenging.
• Automating car-sharing fleet relocation in cities presents significant decision-making hurdles, particularly for platooning and parking maneuvers.

Purpose of the Study:

• To propose a novel behavioral planner framework for automated car-sharing relocation in urban settings.
• To address the decision-making challenges in managing platooning, parking, and de-parking maneuvers for urban car-sharing fleets.

Main Methods:

• Developed a Finite State Machine (FSM) based behavioral planner.
• Incorporated four key maneuvers: platoon following, parking, de-parking, and platoon joining.
• Implemented a Vehicle-to-Vehicle (V2V) communication protocol for platoon management.
• Utilized classical (PID) and Model-based Predictive Control (MPC) for longitudinal and lateral vehicle control.
• Validated the planner in a simulated urban environment using the Carla Simulator.

Main Results:

• The proposed behavioral planner successfully managed automated car-sharing relocation maneuvers in a simulated urban scenario.
• Demonstrated the framework's capability to handle complex urban driving conditions and decision-making.
• Validated the effectiveness of the integrated V2V communication and control strategies.

Conclusions:

• The novel FSM-based behavioral planner offers a viable solution for the automated relocation of car-sharing fleets in urban environments.
• The approach effectively integrates decision-making, communication, and control for autonomous urban mobility challenges.
• This research contributes to advancing the practical implementation of autonomous systems in shared urban transportation.