IMPROVED METHOD FOR ANALYZING TRAVEL TIME RELIABILITY

摘要

An improved method for analyzing travel time reliability. On the basis of traditional travel time reliability index estimation, a Bonferroni-corrected K-S detection method is used to test whether seasonal or monthly differences exist in travel times; log-normal distribution is used to fit the distribution of travel times, and an expectation maximization algorithm is used to perform parameter calibration; the Bootstrap method is used to perform sampling with replacement on travel time estimation samples; M rounds of resampling are performed; a travel time reliability index TTRm is calculated for each data subset; and point estimation and interval estimation are performed on TTRs that accord with a normal distribution. The method outputs a reliability index estimation interval, the confidence interval of which, compared to that of a reliability index point estimation value, more comprehensively describes travel time reliability. The upper limit of the interval is also more practically significant in applications such as guidance issuance and traffic planning.