Autonomous vehicles operate in real time traffic scenarios and aim to reach their destination from their source in the most efficient manner possible. Research in mobile robotics provides a variety of sophisticated means with which to plan the path of these vehicles. Conversely professional human drivers usually drive using instinctive means, which enables them to reach their goal almost optimally whilst still obeying all traffic laws. In this paper we propose the use of fuzzy logic for novel motion planning. The planner is generated using an evolutionary algorithm which resembles the learning stage of professional drivers. Whether to overtake or not, is a decision which affects one’s driving and the decision is made using some deliberation. We further extend the approach to perform decision making regarding overtaking for all vehicles. Further we coordinate the motion of the vehicles at a traffic crossing to avoid any potential jam or collision. Experimental results prove that by using this approach we have been able to make the vehicles move in an optimal manner in a variety of scenarios.
Bibliographical noteAvailable under Open Access
- autonomous vehicles
- motion planning
- fuzzy logic
- evolutionary fuzzy inference systems
- traffic simulation
- multi-robot motion planning