A Novel Self-Organizing Approach to Automatic Traffic Light Management System for Road Traffic Network

Abstract

Traffic network is basically a “network of networks” consisting of mainly two types of networks: road network and a travel network. Due to drastic increase in population of vehicles, traffic congestion in metro cities of India is a severe problem. To resolve this use, we have proposed an approach to dynamic and automatic road traffic light management system. The approach is based on arithmetic mean theorem. The input to the control interface of traffic light of proposed approach are six different parameters : velocity of traffic units (v), queue length (l), inter-arrival time between vehicles (t), centrality measures value(c) and predicted value (prdt v) of traffic congestion by historical database and output is level of congestion. On the basis of congestion level, we build inference rules. The proposed approach automatically updates duration of green and red light as per the level of congestion at a particular junction. Proposed approach follows forward reasoning of If–Then rules. A study of this approach is done in this paper on various traffic situations of Delhi, India depending on queue length of vehicles.