抄録
Simulated Annealing (SA) is an effective general heuristic method for solving many combinatorial optimization problems. But SA has two problems which are the long computational time of the numerical annealings and the determination of the appropriate temperature schedule. For the temperature schedule, Harry found that a specific constant temperature in SA yields good solutions for TSPs. The number of such specific constant or important temperature region has been thought to be one. But, Miki found that there are two important temperature regions in some TSPs. This paper discusses the multiple important temperature regions, and provides a guideline for determining proper temperature schedules in SA.
