This study is originated with the intent of qualifying traffic string stability from empirical observations. A new responsiveness angle measure is developed to assess the driver's instincts under vehicle-following conditions. In this measure, the degree of the follower vehicle's attention towards its leader vehicle's actions is quantified. In understanding the string stability in the traffic stream and assess the propagation of disturbances, the newly conceptualized measure is used along with a discrete Fourier transform to measure the frequencies associated with the responsiveness angle sequences. In this transform, a higher frequency of the angle depicts unstable conditions and vice versa. In assessing string stability from the empirical observations, vehicular trajectory data were developed from three study sections. Two study sections tend to have homogeneous lane wise traffic, while the third section had mixed (heterogeneous) traffic. The results of the string stability analysis over the study sections showed that string stability varied with the change in traffic flow conditions, road geometries, and traffic flow type. In the case of free-flow conditions, the traffic streams were found to be stable with marginal disturbances in the responsiveness angle. From the analysis, it is observed that, in the case of study section 3, around 26 instances of the stream are extremely unstable conditions (frequency is equal to 10). For study section 1 and section 2, the traffic stream is unsteady for 4 and 13 instances, respectively. However, as the traffic flow level rises, string stability deteriorated. This study demonstrated a novel approach to analyze string stability based on actual traffic conditions that can be implemented in real-time for traffic stream monitoring.
|Journal||ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Part A: Civil Engineering|
|Publication status||Accepted/In press - 2022|