Multiple-objective simulated annealing optimization approach for vehicle management in Personal Rapid Transit systems

Abstract

This paper presents a multi-objective simulated annealing (MOSA) algorithm for the static problem of routing electric vehicles with limited battery capacity in a Personal Rapid Transit (PRT) system. The problem studied in this work aims to minimize both the total energy consumption and the number of vehicles used. Our algorithm uses a strategy of Pareto-dominant-based fitness to accept new solutions. The performance and computational costs of MOSA are studied on a set of randomly generated instances. Algorithm is found to be effective for the multi-objective version of the PRT problem.