Interrupted Flow Link Travel Time Estimation Based on Video Detector Data and Microwave Detector Data

Abstract

Due to the fact that penetration rates of both video detectors and micro-wave detectors are quite low on interrupted flow, it is very hard to effectively estimate link travel time on the whole road network based on these two detector data alone. In order to solve the problem, a data fusion model is being brought forward to estimate link travel time based on video detector data and microwave detector data. In the model, link travel time consists of two parts--travel time without signal delay and signal delay, microwave detector data is used to estimate travel time on the link without signal delay and video detector data is utilized to estimate signal delay at the downstream intersection. In the end, field data from Suzhou is collected to validate the model, and results prove to be satisfactory.