A methodology for estimation of travel time statistics on urban network

This paper presents a methodology that integrates cumulative plots with probe vehicle data for estimation of travel time statistics (average, quartile) on urban network. The integration reduces the relative deviations amongst the cumulative plots so that the classical analytical procedure of defining the area between the plots as total travel time can be applied. For quartile estimation, a slicing technique is proposed. The methodology is validated with real data from Lucerne, Switzerland and it is concluded that the travel time estimates from the proposed methodology are statistically equivalent to the observed values.