A fast optimization method based on a hierarchical strategy for the travelling salesman problem

An algorithm based on a hierarchical strategy is proposed for solving the travelling salesman problem. The tour patterns have been analyzed and compared with those generated from different stages of simulated annealing. A set of problems with known solutions has been used to evaluate the new algorithm. The hierarchical algorithm is shown to be very fast. It finds a path length similar to that obtained using simulated annealing algorithm. The path length is quite far from the global minimum but compares favorably to that obtained using simulated annealing.