Abstract

The main objective of this study was to implement a computational prototype in two stages. The first stage focused primarily on generating efficient routes based on an evolutionary algorithm, while the second stage focused on determining the fleet and frequencies based on an assignment algorithm, in order to tackle the complex combinatorial search problem of a public transportation network. The prototype utilizes genetic algorithms (GA) as the metaheuristic in the first stage and an assignment algorithm using the routes generated in the first stage in the second stage. The results show that in the first stage, the GA metaheuristic consistently provides highly efficient routes in each execution, demonstrating that the combinatorial complexity of the problem is resolved in the first stage. The results were validated on the Swiss Road network by Mandl, and they proved to be better solutions than those presented in previous studies.