魔方求解问题的设计与实现.docVIP

魔方求解问题的设计与实现 摘要Abstract The thesis introduces a program for quickly resolving the Rubik’s Cube problem. By exploiting the methods in the expert system theory, the program can give the rotation sequence for any inputted state to the destination state. The program is a part of the whole program for automatically solving the Rubik Cube and demonstrating it in animation. The core of the program is a module of expert knowledge, a representation of rotation sequences. The program repeatedly matches the current state with the states in the representation. If a math happens, the corresponding rotation operation will be called and a new state will be obtained, so that the destination state can be found at last within a short time. The program exploits the A* algorithm for graph searching. It is a common algorithm based on node expansion, but it does not define the actual expanding method of search trees. Here, the breadth-first search is used so that the best path may be found. The program integrates the breadth-first search with the A* algorithm to traverse the state graph and conduct node expansion. In programming, backtracking is employed in the process of searching. Sometime the program cannot give the answer to some special inputted state. It is not because of the algorithm which the program using, but because of the limitation of the expert knowledge Key　words：rubik cube; backtracking; best path 目录第1章　概述 1 1.1魔方问题简介 1 1.2　关于搜索与存储的问题 1 1.3　关于专家系统的简介 2 1.4　文件的输入 2 1.5　本章小结 3 第2章　程序总规划 4 2.1　程序的总体设计 4 2.2　全局变量的定义 5 2.3　各个模块间的调用关系 5 2.4　本章小结 7 第3章　各个模块的详细设计 8 3.1　主模块 8 3.2　文件接收模块 9 3.3　程序执行模块 10 3.4　旋转操作模块 12 3.5　搜索模块 14 3.6　专家系统模块 18 3.7　本章小结 20 第4章　程序演示 21 结　论 24 参考文献 25 致　谢 26 第1章　概述 1.1魔方问题简介 魔方（RUBIKS CUBE）是由匈牙利的厄尔诺·鲁毕克在1979年设计的。他是一个建筑学家，在布达佩斯执教。同时日本的石毛也独立完成了这个设计。两人都取得专利。设计魔方的目的在于帮助人们更好的理解三维空间里的各种运动。 魔方是一个看似简单的玩具，它的每个面有九个格。我们目标是使每个面的格子颜色都相同。每个格子实际上是小方块的一部分。小方块的每个面都可以旋转，角方块有三个面，边方块有两个面。一共有八个角块，十二个边块和六个中块。 目标是把一个每面都是随机颜色的魔方还原成每面颜色都相同的魔方。问题是，有无数的方法操作魔方，然而只有极少的方法能够