A Personalizable Driver Steering Model Capable of Predicting Driver Behaviors in Vehicle Collision Avoidance Maneuvers

Abstract

In recent years, significant emphases and efforts have been placed on developing and implementing advanced driver assistance systems (ADAS). These systems need to work with human drivers to increase vehicle occupant safety, control, and performance in both ordinary and emergency driving situations. To aid such cooperation between human drivers and ADAS, driver models are necessary to replicate and predict human driving behaviors and distinguish among different drivers. This paper presents a combined driver model that is able to not only identify different individual driver behaviors, but also predict a driver's behavior in rare vehicle maneuvers such as collision avoidance (CA) based on his/her daily driving data. The driver model consists of a compensatory transfer function and an anticipatory component and is integrated with the design of the individual driver's desired path. It has been shown that the proposed driver model can replicate each driver's steering wheel angle signal for a variety of highway and in-city maneuvers. The utility of the proposed driver model is its ability to predict a driver's steering wheel angle signal for a CA maneuver from only daily nonemergency driving data. The driver model is then validated by comparing two different drivers’ model parameter sets to the group average to show that each driver has a unique set of parameters. Finally, the driver model is validated by showing that its daily driving parameters differ from its predicted CA parameters.