Dynamic traffic prediction for insufficient data roadways via automatic control theories

This study aims to propose an algorithm for traffic estimation, particularly for urban roadway sections with insufficient data sources. Establishing an advanced traffic management system (ATMS) heavily depends on obtaining complete data. In this regard, traffic detectors play an important role in data acquisition. However, installing traffic detectors extensively over a metropolitan area can be rather expensive. As such, in most cities around the world, vehicle detectors for traffic data collection are relatively insufficient to meet requirements. In light of this, the current research therefore proposes a feasible algorithm of dynamic section flow prediction based on the automatic control theory and observer design for roadway sections with insufficient data sources. According to the numerical evaluation, the method is found to be satisfactory and acceptable in practice.