Dynamic State Estimation of Driving Style Based on Driving Risk Feature

Abstract

In this paper, a novel method is proposed to estimate the driving styles of other drivers based on a driving risk feature. This new feature is proposed using a dynamic potential field method wherein the distribution changes depending on the relative number of adjacent vehicles. A more appropriate description of driving risk is obtained compared to other indices. The proposed feature dramatically improves the accuracy of estimating the driving style. To estimate the driving styles, this study considers a problem of the fundamental model under a scenario wherein the target vehicle follows the preceding vehicle. The proposed estimation method is validated through experimental results