The current paper outlines a comprehensive methodology for modeling speed data on two-lane roads during periods of heavy traffic characterized by heterogeneous vehicle types. In such scenarios, the presence of diverse vehicle types causes a significant departure from the normal speed distribution model. This deviation becomes particularly pronounced during heavy traffic flows due to the frequent interactions among vehicles within the traffic stream. Consequently, there arises a necessity to develop a modeling approach specifically tailored to such flow conditions. Drawing upon empirical data collected from a major intercity road in India, this study uncovers a notable skewness in speed data under heavy traffic conditions. This skewness primarily stems from the formation of vehicle platoons within the traffic stream, exerting a substantial influence on their speed characteristics. By scrutinizing the distribution of this data, the study concludes that a logarithmic transformation effectively aligns with the assumption of normality. This assertion is supported by various goodness-of-fit metrics, affirming the suitability of the proposed modeling approach for capturing the intricacies of speed behavior in heterogeneous traffic environments.
Read full abstract