象棋殘局庫之研究

Abstract

回溯分析演算法已經應用在許多棋類AI程式，中國象棋因為棋規問題無法直接套用傳統的回溯分析演算法，最近幾年方浩任等人研究出建立包含亞洲棋規的象棋殘局庫，使象棋殘局庫變的完整。為了提升象棋AI程式的棋力，需要花費的大量的時間與空間來建立象棋殘局庫。

本篇論文研究象棋殘局庫，利用象棋的兵卒走法的特性分割象棋殘局庫，並提出使用映射法進一步壓縮象棋殘局庫，最後評估不同方法對於象棋殘局庫的建立時間與使用空間的影響。

Chinese Chess endgame databases including Asia Rule were constructed several years ago. In order to enhance Chinese Chess AI program at the endgame stage, it needs to spend a lot of time and space to create Chinese Chess endgame databases.

This paper uses the characteristic of pawn of Chinese Chess to split Chinese Chess endgame databases and then use a mapping technique to compress them. And evaluate the impact of the time and space cost in different methods for the construction of them.