Deducing and forecasting expressway status based on toll collection data

Expressway transportation is playing a more important roll in economic construction and social life. The expressway network status, including section flow and average speed, is important to the expressway management for analysis and decision making. Traditional ways of monitoring the road is using detectors which cost much to install and maintain. We propose speed constraint deducing model and similarity weighted forecasting model to calculate expressway network status based on toll collection data. The deducing model calculates all the cars´ movement in the network according to the time and stations they entered and exited the network, and then generates road status of the past time. The forecasting model forecasts the current and shot term future status based on weighted historical status, when toll collection data is not available. Experiments based on toll collection data of Beijing expressway showed that the deducing model and forecasting model had good accuracy. The models present network status without any cost, which have a broad application prospect.