A*（エースター）と言われるアルゴリズムが一番よさそうだが
１マス移動毎にやってると重いし
マップに入って１回しかしないと環境の変化（別NPCの挙動）に対応しにくいし・・・
ターン制のシミュレーションRPGとかだとターンのタイミングでやるんだろうけど
ゴールする事以外の別のトリガーで発動させないといけないな・・・
さてさて、どうしたものか・・・
余談ですが、私はスタンリーキューブリックが大好きです。
------------------------
[Translation with a translation engine]
It is the place which considers A.I. Artificial Intelligence of the character which walks on a map.
The algorithm called A* (A-Star) seems to be the best.
It is a pile when it is doing for every 1 mass movement.
It goes into a map, it turns one and lends, and if there is nothing, it is hard to respond to an environmental change (action of NPC).
In the simulation RPG of a turn system, I think that it performs in the timing of a turn change.
You have to make it move by another trigger except "making a goal."
Now, it would not do what...
Although it is a digression, I love Stanley Kubrick.
