A New Online Travel Time Estimation Approach using Distorted Automatic Vehicle Identification Data

Online travel time estimation is an important procedure for real-time traffic information systems (RTIS). In this paper, we describe a preliminary travel time data collection and estimation platform developed for RTIS application based on automated vehicle identification technique deployed in the Stockholm city area. The platform is composed of a client-side travel time analysis program and a database server. To obtain accurate real-time link travel times for traffic state prediction and RTIS applications, an optimal filtering algorithm is developed and evaluated using travel time data collected on urban streets in and near the city of Stockholm. The proposed algorithm shows reliable performance against the highly noisy traffic context, and is more robust than existing online travel time estimation algorithms. The estimated travel time information provides a solid basis for advanced traffic information system applications.