Urban Area Vehicle Number Estimation Based on RTMS Data

Along with the increase of vehicle ownership, the traffic problem has a serious impact on people's daily life. Not only the traffic congestion, but also the parking problem troubles urban daily traveling. Therefore it is important to obtain the parking demand to help the government to make a rational decision on traffic planning and management. This paper focuses on estimating the vehicle number in a certain area (i.e., the spaces surrounded by the arterial roads) in each time slot to analyze the area parking demand, using RTMS (Remote Traffic Microwave Sensor) data. We first propose a basic method to calculate the AVN (Area Vehicle Number) based on the inflow and outflow traffic of the area. In order to correct the error caused by minor roads without RTMS data, we propose an advanced method to improve the estimation accuracy by exploiting the road traffic correlation from a network perspective. Comprehensive evaluation is conducted to verify our design based on large amount of RTMS data from the Hangzhou city during one month. The estimation results also demonstrate interesting human behaviors among various urban areas.