Parameter Tuning in MACO for Actual Road Conditions

Abstract

With the increase in traffic volume day by day, cities are facing the problem of extreme congestion in most of the developing countries. The congestion on the road leads to increase in travel time and travel cost as well as having a significant impact on the health of people. This paper aims to provide the parameter tuning in the existing modified ant colony optimization (MACO) algorithm using particle swarm optimization (PSO) algorithm to handle the congestion for actual road conditions. The performance of the MACO algorithm depends on the value of its parameters. Having a very large combination of these parameter values, a selection of the optimal values for these parameters is done using trial and error method. Therefore, we have proposed an algorithm based on PSO algorithm to find an optimal combination of the parameters of MACO under actual road conditions. MACO algorithm works under the assumption that all roads are in working condition, whereas the proposed work finds the best path even when the roads are temporarily blocked may be due to an accident or permanently blocked due to road construction. The experiments have been done for the network of North-West Delhi, India. It was found that the travel time is reduced significantly after using the optimal combination of values of parameters.