Data Mining Based Research on Urban Tide Traffic Problem

Nowadays in some cities, because of inappropriate of layouts of living areas and working areas, every morning, millions of vehicles flood into working areas from living areas, while every evening those vehicles back to living areas, which forms so-called traffic tide phenomenon (TTP) in which vehicles are congested in one direction while the opposite direction is relatively free, especially in some main roads. So in this paper, based on Shanghai outlines cross-river tunnel (SOCRT) project, a Data Mining based traffic direction control algorithm (DMTDCA) is proposed to adjust the traffic direction of Direction-Changeable Lanes (DCLs) in the tunnel automatically and timely according to analysis results of current traffic flow and short-term forecasted traffic flow of two tunnel entrances in order to make full use of all lanes. Field tests and user reports show efficiency of DMTDCA by 30% increase of average traffic capability, 10% increase of rush hour traffic capability and 40% decrease of average queue length.