Computer chess is always an important research area in artificial intelligence. At present, there is less paper dealing with the playing strategies of Dark chess. Dark chess is an incomplete information game with probabilities, which is not the same as complete information games, such as chess or Chinese chess. If we use conventional game-tree searching techniques to tackle Dark chess, then the number of branches will be very large because there are lots of moves for both “dark pieces” and “bright pieces”. Hence, it is not easy to improve the strength of the Dark chess program by using the conventional game-tree searching techniques.
This thesis is written to improve the Dark Chess program which was developed by the postgraduate Hsieh,Yao-An. We improve his move generator first, and then the evaluation function. As his evaluation function used static scores to calculate the materials’ values, regardless of how the chess game plays out, in many cases it will lead to wrong judgments. We want to dynamically change the chess materials scores when the chess board is changed to a more objective measurement. We carefully consider the unique food chain relations of the chess species and design a Dark Chess program to enhance the evaluation function. Finally, we combine several techniques to improve the strength of the program.