DRIVER MODELING USING FUZZY LOGIC CONTROLS FOR HUMAN-IN-THE-LOOP VEHICLE SIMULATIONS

ABSTRACT It is a universal trend that automotive engineers tend to use fast digital computers to develop advanced vehicle systems. They can design, analyze, and test their systems using computer simulation before physically manufacturing them. It is still questionable how these advanced systems would react with the human driver. One way to deal with this problem is to develop a computer model that is capable of controlling the vehicle in a way similar to human driver behavior. Fuzzy logic inference systems are known of their great ability to simulate human reasoning process as well as the possibility of being further trained to mimic specific human control process. This paper presents a new driver model using fuzzy logic controls. The model is designed to control the longitudinal as well as the lateral motions of the vehicle by performing simultaneous steering and braking commands. The model is tested on a vehicle model having an integrated active steering and direct yaw control strategy as developed by this paper's authors in [1]. The results show success of this fuzzy model in simulating driver control actions in curve following and collision avoidance maneuvers.