The Design And Implementation Of Gomoku Playing Platform

As a important branch of artificial intelligence(AI), computer game research can even date back to a mechanic automatic chess player of eighteenth in Europe. Over last fifty years, with the advances of computer science, the study of computer game advanced significantly, which contributed a lot to the theory and method of AI. And thus the advances of computer game has been the benchmark of AI. Since AI study began in America, the original subject of AI is chess. Over these years, computer chess has been studied extensively and deeply, with many fruitful wok and results. But the study of gomoku is relatively slow-paced. Though a lot can be borrowed from the results of computer chess and Chinese Chess, gomoku has its unique features. Typically, chess can use 64 bit integer number to represent its pieces since its board is eight multiply eight, but this is not applicable to gomoku. Another feature of gomoku is that it has a huge search space. Therefore, the study of gomoku algorithm engine and the design of gomoku software, and thus implementing a platform of gomoku both for playing and stduding is meaningfull.Based on a broad study of gomoku theory papers and deep understanding of the current development and practice of gomoku game, this thesis aims at design and implementation of a gomoku playing platform, which supports human-machine playing, human-human playing, machine-machine playing and network playing. The system’s design complies with the general protocol of World’s Gomoku Cup, making it easy to import 3rd-party engines; with the help of QT library, it support Windows, Mac and Linux platform; By utilizing socket protocol, it support network playing. The platform not only serves as playing platform, but also as researching platform of gomoku. It accomplishes the following work:1. Study the components of computer gomoku game such as board representation search algorithm, move generation, and evaluation function.2. Study optimization method based on Alpah-Beta search algorithm. Design new data structures and board representation to accelerate search algorithm; give a improved algorithm base on permutation table.3. Study the board knowledge base and pregenerated system, which can be used to make the search algorithm more efficient.4. Apart from studying the gomoku AI engine, this thesis designs a GUI platform which supports human-machine playing, human-human playing and etc. The platform also supports loading chess manuals and saving game information, so it can be used as a researching platform.