Link Average Speed of Traffic Flow Estimation Method Based on Floating Car

Abstract

In floating car data collection, the link average speed of floating cars is exported as link average speed of traffic flow. This results in low quality traffic information being shared with the public. This paper investigates the quantitative relationship between link average speed of floating cars and link average speed of traffic flow considering the traffic conditions, and verifies the research with simulated data based on a part road network in Changchun. The results show that the link average speed of floating cars can not be used instead of the link average speed of traffic flow, and that the link average speed of traffic flow can be estimated accurately using the method described in this paper. INTRODUCTION Link average speed is regarded as an effective traffic parameter for measuring the degree of traffic congestion. Floating car data collection has become one of the main means for collecting link average speed (Jiang G., 2004). In typical case, a single vehicle type (taxi, bus, logistics vehicle etc.) is selected as the floating car in order to reduce system costs (Turner S.M., 1998). However, the structure of vehicle type in traffic flow is very complex, and the running characteristics of different types on road are distinctive. Therefore, the Link Average Speed of Floating Cars (LASFC) cannot usually be substituted for the Link Average Speed of Traffic Flow (LASTF). Currently, research on floating car-based traffic information collection technology focuses on traffic data estimation methods based on individual floating car, and the average traffic data of floating cars is directly exported to users (Quiroga, C.A., 1997, Young-Ji, B., 2005 and Dong, J., 2006). The relationship between traffic data for floating cars and traffic flow is infrequently discussed. ICCTP 2009: Critical Issues in Transportation Systems Planning, Development, and Management ©2009 ASCE 1631