Institute of Transportation, MOTC

       When the traffic demand is greater than the capacity provided by the road, the vehicle speed will drop sharply and enter a congestion state, and this phenomenon is called traffic breakdown. The queue discharge rate after the traffic breakdown is generally lower than the pre-breakdown flow, which is a capacity drop. This phenomenon can lead to negative effects such as a decrease in the road capacity, an increase in travel time and the risk of accidents. Capacity drop can be discussed from macroscopic and microscopic perspectives. Early studies used vehicle detector data in the analysis. With the advancement of computer technology and the acquisition of traffic trajectory data in recent years, the behavior and characteristics of individual vehicles can be inspected from a microscopic perspective. In the previous studies, the reasons for capacity drop were still not clear, and the perspectives for discussing this topic were also different. From a macroscopic perspective, factors such as traffic flow characteristics, traffic management methods, and road geometry configuration at the macroscopic level have been investigated. From a microscopic perspective, traffic oscillation and lane changes behavior were found to be the most related factors. In this regard, different scholars have different views on capacity drop, so there was no unified conclusion on capacity drop. In addition, most of the past studies that analyzed from a microscopic perspective used the trajectory database data of NGSIM in the United States. Due to the limited time range of data collection, the issue of capacity drop during the stable and discharge traffic flow has not been fully discussed. In addition, some scholars have found that the cause of the traffic oscillation is the behavior of lane changes. Still, other scholars believe that even if there is no lane-changing behavior, the traffic oscillation will be caused by the unstable car-following behavior. The same driving characteristics are not consistently maintained on the road, so the impact of the type of driving response on traffic also needs to be considered. Here, this study uses the trajectory data of the on-ramp merge areas of Taiwan freeways. The time range covers from the free flow state to the congested state. In addition to estimating the capacity drop of Taiwan freeways, it will also target the traffic oscillation during the period of traffic stabilization discharge. Discuss the influence of aggressive and timid driving behavior on the discharge rate.
This study also explores the magnitude of the capacity drop at the macroscopic level and the factors that affect the capacity drop at the microscopic level. First, used the VD data to check the distribution of capacity drop in the time dimension, and confirm that the capacity drop is a fluctuating value that has little correlation with road geometry, but may be more related to driving behavior. Then, used the traffic trajectory data check from the spatial dimension. It is found that the traffic breakdown does not necessarily occur in the entire section of the lane and individual lanes at the same time, which may be one of the factors that caused no unified conclusion on the issue of capacity drop in the past; in addition, the critical point on the road is located in the range from the end of the acceleration lane to the 80 meters downstream of the on-ramp merge areas. After confirming the location, using the trajectory data to estimate the magnitude of the capacity drop, the rate of the National Highway No. 1 Touwu Interchange is 19.34%, the National Highway No. 3 Zhonghe Interchange is 3.01%, and the National Highway No. 3 Xindian Interchange is 10.51%. Finally, the impact of traffic oscillations occurring during traffic stabilization discharge period on capacity drop is discussed. The oscillation cycle can be divided into a precursor stage, a growth stage, a stable stage, and a decay stage. The more conservative Concave and Convex are the majority of the reaction types, which will lead to a decrease in the discharge rate in the precursor stage, growth stage, stable stage, while the more aggressive Constant and Other are the majority in the decay stage, which will lead to an increase in the discharge rate. It is hoped that through the results of this research, it will provide a reference for future academic circles to discuss the issue of capacity drop and practical application units in the development of traffic management strategies.