Estimation of Driving Performance Level Using Longitudinal and Lateral Driver Models

This study proposes model-based driver performance level estimation method. Assuming that driver performance level in drowsiness driving state can be potentially detected from driving behavior deviated from driver's normative characteristics. The normative characteristics are estimated by employing driver models considering the transfer function from driving environment to driving operation. Therefore, the driver models are constructed using only driving data which can be measured with existing on-board vehicle sensors. In addition, to estimate the driver performance level by the criteria of the individual driver, the driver assistance system can be adapted not only to various types of drivers, but also to real-time driving performance level.