Koetsu OMINE

Analysis of Factors Affecting Right-Turn Gap Acceptance Behavior and Model Development

Kazushi SANO

This study focuses on right-turn behavior, which becomes complex due to strong influences from oncoming vehicles, aiming to quantify its decision-making structure. The characteristics and variability of right-turn gap acceptance at intersections were analyzed, and a right-turn decision model was constructed to capture fluctuations in capacity across intersections and time periods.
First, the critical gap was used to evaluate factors affecting right-turn behavior, with analyses under various conditions, such as by intersection and weather. The results confirmed that required gaps differ by intersection layout and vary even at the same intersection depending on time of day and traffic volume. Differences in driving conditions, such as lane and type of oncoming vehicles, also affect required gaps, and adverse weather, such as snowfall, significantly increases them. These effects were quantified, showing that right-turn decisions are influenced by intersection structure, oncoming traffic, and weather conditions.
Next, considering these factors, a right-turn decision model was developed for each intersection. Results showed explanatory variables differed by intersection, suggesting that structural conditions and visibility influence decision factors. On the other hand, the gap and speed of the vehicle following the oncoming vehicle were consistently selected across all intersections, confirming them as primary factors.
To account for multiple vehicles turning consecutively within a single gap, a sequential logit model was developed to replicate stepwise decision-making for the first, second, and third right-turning vehicles, predicting the number of vehicles turning. Results indicated that approximately a 2-second increase in the gap is needed for each additional vehicle. Using this model to estimate vehicles able to turn during a green signal revealed that right-turn capacity decreases significantly during the morning peak, highlighting temporal fluctuations. Differences in signal control and oncoming traffic characteristics were also found to influence capacity.