Swarm simulated annealing algorithm with knowledge-based sampling for travelling salesman problem

Simulated annealing (SA) algorithm is a popular intelligent optimisation algorithm, but its efficiency is unsatisfactory. To improve its efficiency, this paper presents a swarm SA (SSA) algorithm by exploiting the learned knowledge from searching history. In SSA, a swarm of individuals run SA algorithm collaboratively. Inspired by ant colony optimisation (ACO) algorithm, SSA stores knowledge in construction graph and uses the solution component selection scheme of ACO algorithm to generate candidate solutions. Candidate list with bounded length is used to speed up SSA. The effect of knowledge-based sampling is verified on benchmark travelling salesman problems. Comparison studies show that SSA algorithm has promising performance in terms of convergence speed and solution accuracy.