Multi-Modal Routing Using NSGA-II Algorithm considering Covid-19 Protocols: A Case Study in Tehran

Document Type : Original Article


1 Geomatics and Geodesy, GIS, K. N. Toosi university, Tehran, Iran

2 Geomatics and Geodesy, GIS, K. N. Toosi University, Tehran, Iran

3 University of Tabriz, Tabriz, Iran



Nowadays, the utilization of urban public transportation has become a routine for people living in large cities. It is necessary to use multi-modal along the path to move individuals from one point to another. With the initiation of the Covid-19 pandemic, megalopolis public transportation became an effective place for transmitting the virus. Hence, this study used a multi-modal route according to health protocols. Tehran is home to the largest transportation network in Iran, so its public transportation network has been used in this paper. Subway, bus, taxi, and pedestrianism are 4 investigated structures. Also, the multi-modal routing problem is solved as multi-objective and by using the NSGA-II algorithm. The effective goals in choosing an optimal path according to the health protocols of Covid-19 include traffic, path length, travelers' contact with surfaces, air conditioning, and social distance. A basic Genetic algorithm was used to verify the selected path by the NSGA-II algorithm and a new algorithm that has been improved base on NSGA-II in this paper which has been named MPNSGA-II(Multi-Parent NSGA-II): Five target functions were evaluated one by one using the Genetic algorithm, and their results were compared with those of the NSGA-II, and MPNSGA-II algorithms. Based on real data obtained from public transportation in Tehran, the proposed routes are rational and acceptable. GA could solve in 4.4 seconds and NSGA-II and MPNSGA-II in about 10 seconds. MPNSGA-II generally can reach a better amount of variance, for example, the Distance between people's objective function for MPNSGA-II is 27.34, for NSGA-II is 62.21 and for GA is 48.73.