A scalable linear algorithm of simulated annealing (SA) was proposed by Lin et al. that is capable of achieving a near optimal solution for the travelling saleman problem in a controllable way. Since the linearity is based on the hybrid mechanism that combines SA heuristics with the scaling relation of acceptance ratio in the low temperature, other conventional heuristics in optimisation problems ought to be tried. The nearest-neighbour (NN) heuristics is thus studied, and one finds that the quenched configuration of NN's could be resurrected back to SA path by the hybrid mechanism. It is also verified that the same scalable linear algorithm of Lin's may continue to apply with exactly the same set of parameters.
關聯:
Physica A: Statistical Mechanics and its Applications 203(3-4), pp.369-380
