Development of a realistic driving behavior by means of fuzzy inference system

Abstract

Realistic traffic flow simulation is possible when the vehicles inside the simulation are able to mimic human driving behavior. In view of this, this paper will discuss the implementation of fuzzy logic inside the Behavior Model framework with the intention to develop intelligent simulated vehicles. This Behavior Model consists of three different units, namely; Vision and Perception, Decision and Motion Control Unit. Vision and Perception Unit acts as the eyes for the intelligent vehicle. Decision Unit will decide the maneuvering decision. Finally, Motion Control Unit will transfer the decision into motion. However, the implementation of fuzzy logic with the integration of fuzzy rules and defuzzification techniques is done in the first and second units. This Behavior Model is controlled by two sets of fuzzy inference systems (FIS) which are free flow vehicles following and changing lanes. The finding of this research shows that the Behavior Model with fuzzy logic is able to create an intelligent vehicle that is able to self-maneuveri inside the traffic flows, realistically.