Prediction of Driver's Brake Pedal Operation in Vehicle Platoon System: Model Development and Proposal

摘要

This paper focused on the remaining time to brake pedal operation (TTBP) and proposed the unscented Kalman filter (UKF)- and particle filter (PF)-based models to estimate the TTBP prior to the actual brake pedal operation. The state and measurement equations were simply modeled by partially using an ANN model. Also, the three-vehicle convoy data collected through a driving simulator (DS) experiment showed that the first car started decelerating and shortened the headway two to three seconds before the brake pedal operation. This made sure that the proposed UKF and PF could be applied to the estimation of TTBP by simply measuring the velocity and headway of the preceding cars.