Benchmarking Metaheuristic Optimization Algorithms on Travelling Salesman Problems

A wide variety of metaheuristic optimization algorithms has been proposed to optimize the travelling salesman problem (TSP), and methods drawing inspiration from nature have been rising in popularity due to their competitive performance in regard to quality of solution and computational efficiency. In this paper, a selection of algorithms, using standardized parameters frequently found in literature, is applied to a multitude of TSPs with differing complexities and sizes, and subsequently benchmarked on solution quality and convergence. We consider the quality of the implemented algorithms using standardized and not a specifically optimized parameter choices according to a certain problem.