Stability Analysis Based on Instantaneous Driving Behavior Using Car-Following Data

An attempt is made to analyze the stability of a platoon using experimental data measured by real-time kinematic (RTK) Global Positioning System (GPS) receivers. Car-following experiments were conducted on a test track using 10 passenger cars. Various speed patterns were tested for the lead car, including random, constant, and sinusoidal, giving different driving conditions. The responses of the following drivers were measured by RTK GPS receivers in each car. The stimulus–response car-following concept was examined, assuming that the reaction time might vary over time. A graphical method was modified to estimate the time-variant reaction time more efficiently. A new algorithm was proposed to estimate the sensitivity factor using Lissajou’s diagram between relative speed and acceleration. The statistical analysis showed that intrapersonal variability was higher than interpersonal variability for both reaction time and sensitivity factor. However, the influence of the driver’s position in the platoon and speed patterns was low. It was found that the reaction time was distributed in a lognormal function for most of the drivers. The variations in the estimated values for the sensitivity factor were relatively high. The stability analysis showed that the average responses of drivers were unstable both locally and asymptotically. The influence of speed fluctuation frequency was found insignificant for the stability of the platoon.