Irregularly Sampled Transit Vehicles Used as Traffic Sensors

Abstract

Performance monitoring is an issue of growing concern both nationally and in the state of Washington. Travel times and speeds always have been of interest to traveler information researchers, but there is limited infrastructure with which to collect such data on a continuous basis. Transit vehicles were used as probes, and a framework was developed for modeling the time series that arise from irregularly sampled transit vehicle locations. These samples of vehicle location were obtained from the King County Department of Metropolitan Services automatic vehicle location system. An optimal filter method that estimates speed as a function of space and time was developed. An optimal solution for the state vector, containing the variables speed and position, was made at each time step by using a Kalman filter.