Abstract
Many researches have analyzed the characteristics of car traffic flow. However, there are limited studies focusing on mixed traffic flow which is very common in developing countries. In a mixed traffic flow, the behavior of cars can be considered as lane-based movement. But motorcycles have non-lane movement because they change their direction and speed very often. Therefore, the paper aims to investigate the mechanisms of interactions between motorcycles in their non-lane movement. This study plan to develop a new approach to describe dynamic movement of motorcycles based on social force approach that is commonly applied to pedestrians. Microscopic data collected from video images taken in a road intersection of Hanoi, Vietnam was used for model validation. Parameters were estimated by using SPSS software with non-linear regression method. The result shows that proposed approach can make reasonable prediction of the histogram choice of speed and direction of dynamic movement.
