Home Index About Next Previous Comments

Effects of vehicle classes and lanes positions on microscopic traffic flow Il-Chul Moon Abstract    Though there are many traffic simulations in the transportation research community, some small features are commonly not represented. Therefore, we perform the naturalistic observation by using a web cam. From the observation, we discover small but important features, such as lane position effects, vehicle type effects and the effects of vehicle types at the first position in a lane. Introduction    Microscopic traffic simulations have been applied to many real world traffic problems. However, some of the features, such as vehicle types or lane positions, are commonly not presented in most simulations. In this paper, we test the relevance of the features by the data from naturalistic observation of a crossroad in Japan. Methods    We setup three hypotheses: the relevance of vehicle type, lane position, vehicle type at the first position in a lane to a traffic load of a crossroad. Then, we perform t-tests on the data collected from the observation. Results    The three hypothesis cannot be rejected, and we conclude that the three factors are relevant to the traffic flow at the spot. Thus, we conjecture that these features may have to be taken into account in future traffic simulations. Privacy    Traffic flow observation with web-cam is less invasive and pretty effective. We cannot recognize the inside of the vehicles because of the low resolution of the cam, but the resolution was good enough to collect the data we need. We believe that the cam can be used for future traffic control purposes without significant invasion to drivers' privacy. References    Camera: https://www.independis.ca/livecams/japan/utsunomiya Project Paper (PDF)