System Design And Implementation Of Gobang Artificial Intelligence

Not only go-bang game is loved by the majority of players,but popular,for programming enthusiasts who would like to explore artificial intelligence,because the rules are simple and popular.Go-bang's man-machine mode is mainly according to the basic rules and principle of the game.Computers need to know where is the following.In order to get it,we need do following steps;Firstly,according to the situation of disk,we calculate importance(called “score”)of every point respectively,which is reached by a possible movement.Then,if there are multiple maximum,we will go through the whole selected numerical biggest point,and randomly select one pointfrom these.One that computer randomly select the biggestscore point,can ensure thatthe position of selected point in every game is different,avoiding chess player to repeatedwin by using a computer.In recent years,the game tree algorithm has been widely used in the go-bang.Game with computer is a very important domain in AI,many production gains from it.Game Tree Searching is a significant aspect for Game.In the world,there are many algorithms to improve the effect of Game Tree Searching,but in China the most used algorithm is also max-mini algorithm.To improve the effect of max-mini algorithm,there are some methods to modify the algorithm with different thoughts.Researching them will help to understand the different methods to improve the Game Tree Searching and master the thoughts of them.Then,with its help a better algorithm will be designed to improve Game Tree Searching.The based on predictive scheduling based on existing alp ha beta pruning and history heuristic based,for extended node,search range is larger,into the finite range limit and the underlying pruning method.At the same time,in order to make the game more interesting and informative,Go-bang game provides two of the way the war.Battle of men and men.In this paper,research work mainly included four parts.Firstly,the basic theories and concepts of the go-bang game areintroduced and researched.Finally,a novel system about go-bang game is represented.Secondly,the paper not only introduces the game tree algorithm,but also summarizes the current game tree optimization methods,including pruning,sorting and history heuristic estimates and give for improvement.Thirdly,the go-bang system is discussed and designed,under the man-machine mode,based on the theory of game tree,this paper gives a valuefor each situation;Then,when the computer chess,it do max-mini search in a certain depth range of chess,find out the best play chess;Finally,alpha beta pruningand limited scope of technology are used to improve speed;Under everyone mode,we add chat function,providing a friendly operation interface,convenient communication of both sides.Finally,the system is implemented by programming and is tested.The go-bang game is completed with Java language,obtains good effect through experimental tests,and can satisfy most needs.