Abstract

This paper presents analysis and modeling of vehicle speed control as demonstrated by human drivers making left turns at intersections. The analysis is based on detailed measurements taken under naturalistic driving conditions. An important application is in the design of future driver assistance systems aimed to reduce intersection collision risk. The model uses the driver's assumed prior estimate of vehicle acceleration based on visual preview information, and unifies estimates of longitudinal acceleration (for braking to rest) and lateral acceleration for negotiating the turn. The relationship between prior acceleration estimates and resulting vehicle accelerations are studied for both stopping and turning events. Closed-loop implementation adopts a nonlinear controller with mode switching functionality. Example simulations show that the model is capable of accurately reproducing measured speed profiles. © 2011 AACC American Automatic Control Council.