Abstract
A generalization of the stochastic method of simulated annealing algorithm based on Tsallis statistics is proposed. This algorithm is considerably faster than the traditional ones in solving the traveling salesman problem. Acceptable solutions are found in fewer steps and higher temperatures than both the classical and the fast simulated annealings. Recent developments in solving NP-complete problems can be incorporated and improve the performance even more.
- Received 16 August 1994
DOI:https://doi.org/10.1103/PhysRevE.51.R1
©1995 American Physical Society