Characteristics Analysis of Road Network Reliability in Beijing Based on the Data Logs from Taxis

Abstract

In this paper, the available data logs collected from taxis are used to analyze the temporal distribution characteristics of the travel time reliability on the road network in Beijing. Three different measures are applied to quantify the travel time reliability during different time-of-the-day (TOD) and day-of-the-week (DOW) periods, namely the Coefficient of Variations (CV), the Planning Rate Index (PRI), and the Probability Indicator (PI). The probability indicator is a novel measure developed in this paper to evaluate the reliability of a road network, which is defined as the probability that the travel rate (travel time per unit distance) is less than a threshold value. The paper then discusses the effectiveness of these measures from the analytical standpoint. The travel rate distribution at an AM peak hour on Wednesday is modeled and the result indicates that a lognormal distribution shows better approximation performance. It is observed that both skewness and width of the travel rate distribution are related to the travel time reliability. Therefore, the probability indicator, considering both width and skewness of the travel rate distribution, is more robust for assessing the travel time reliability than the coefficient of variation and the planning rate index which merely consider the width of the travel rate distribution. Finally, the characteristics analysis of the travel time reliability for different TOD-DOW periods demonstrates that there is a strong relationship between the reliability and congestion. The reliability is low during the peak hours when the congested conditions occur, whereas the reliability is high during the off-peak hours.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 19 citations exist upstream.