抄録
This paper investigates urban expressway capacity characteristics of lane closure sections due to accidents. Methodology includes the following parts: 1) identification of a breakdown due to an accident, 2) estimation of capacity under accident conditions, 3) exploring features of capacity drop percentage, and 4) fitting probability distribution of capacity drop percentage. Accident record data and loop detector data on Odaka Route of Nagoya Expressway during three years are utilized to fulfill the investigation. The results reveal that mean capacity value increases and its variance turns to be convergent with increase of lane closure duration. Accident types like multiple rear-end collisions and rear-end collisions result in comparatively greater values of capacity drop percentage. Also accidents on curve segments bring relatively higher capacity drops compared with those on straight segments. Furthermore, capacity drop percentage is shown to be stochastic, and Beta distribution can well fit its probability distribution.
