Estimation of Travel Time of Different Vehicle Types at Urban Streets Based on Data Fusion of Multisource Data

Due to the limitations of several different kinds of detectors currently in use, typical ways to estimate travel time on urban streets -- based on the data collected by only one source -- are no longer suitable. Moreover, traditional methods of estimating travel time based on all vehicle types on urban streets are unable to provide specific travel times to travelers. This paper aims at providing a model based on data fusion of multisource data to estimate travel times of different vehicle types on urban streets. A weighting calculation that fuses data collected by loop detectors and floating cars is used to build a travel time estimation model for urban road links. The aforementioned fusion method is verified by the simulation software VISSIM. The simulation results lead to the conclusion that this travel time estimation model performs better than the estimation model based on only loop detector or floating car data. The relative error (RE) of travel time is within the range of 10% to 20%. Limitations of the proposed method are also discussed in this paper.