Research And Implementation Of Gobang Game Playing System Based On C/S Model

Artificial Intelligence is one of the active research area in recent years.Artificial Intelligence is a just very strong edge science of arisen comprehensive that develop quickly.Computer game playing is a very popular field in Artificial Intelligence.Research results of Computer game playing contribute to AI not only in theories but in methods as well as remarkable influence to science and social.Computer game playing involves search, reasoning and planning in the artificial intelligence research areas.Gobang's regulation is simpler than Chinese Chess, weichi and Chess in game playing of all kinds.We don't considers an overall situation when we estimate value for some node which function can be very detailed.From the practical angle,in the nowadays Society the leisure network game sets entertainm ent,intelligence - development,interaction and int in a bodyand already become one of the leisure mode mainly of people.The gobang epidemic is broad . Gobang's game regulation is simple , The people of learning gobang is not requiring that very longtime training can be grasped right away.Based on the above consideration, I choose to realize the gobang game playing system to complete the man-machine game playing research subject.The traditional solution game playing question's algorithm is called the search tree method,and also called the game tree method.The computation load is too big if we found all chessgame situation Useing game tree search method, the computer is also unable to withstand.Therefore,if only using the game tree search method to solution man-machine game playing , then the search depth is limited.Judges each kind of move the fit and unfit quality also only to be able to act according to this limited depth to carry on,Even if throughα-βalgorithm , the procedure chess strength level and the search speed is still restricted.In order to solve this problem,it needs us to make the improvement of the search tree method, then enhances the move judgment the accuracy.According to analytical study massive related literature,this paper Combining with the characteristics of Gobang explores a strategy for solving the corresponding game playing problem.Some methods are researched to modify the algorithm,And we give optimized algorithm which is the solution of the gobang game playing,An intelligent gobang game playing system has been developed using JAVA and is realized in the game between computer and human.Search strategy is the key in the game playing.According to the general theory of artificial intelligence and gobang game playing,This paper passes this Gobang game playing program designs,introducing the core contents of the game tree searching and estimate value function.Two aspects were done in the work:1. Gobang's chessgame situation expression method was researched. According to the gobang game regulation, discussing the move is that the board game piece was permitted in the computer.The paper discusses in the artificial intelligence to use in computer game playing the general technical like game tree,maximum and minimum searching,Alpha—Beta Searching.And carries on the estimate value using the static estimate value function to the node.According to the chessgame situation judgment we calculates the estimate value for each legitimate blank node,Then the overall situation value is obtained through synthesis processing.The priority selection is the highest value move.Then calculates the sum of all minute values,and it is the estimate value in the chessgame situation.Quality degree of the static estimate value function decide the height of the intelligent of gobang game playing system directly.If uses the traditional game tree maximum and minimum searching algorithm , computer's decision-making have some flaws. For example,when the computer's or player's three board game piece companies become the straight line, the computer does not attack or the defense Because has the advantageous node which other certain nodes are current win.So,we can Classifies the most main threat and balance threat degree ,give the different value.When has the threat existence, the value of nodes with three or four continual board game piece related is enhanced, The flaw is made up.Because the estimate value function greatly decide the level of computer in playing gobang,I adopt the hereditary algorithm in order to improve the statistic estimate value,And then we could reach the goal of imprving the entire program.The traditional algorithm usually can only maintain group of quite good parameters, while the hereditary mag uphold several move groups at the same time.When need to add a group of new parameters, Normally is to select from the groups of old parameters and combine them together to make a little change.And them compare them with the former parameters. Finally the worst group is removed.Through appraising threaten, the genetic algorithm optimize to the estimate value module. This make the gobanggame playing system have the remarkable enhancement in the chess strength level and the aspect searching of the speed.2.Has completed the C/S pattern client side and the service end design model, gives the related interactive agreement, uses the Socket technology and the multi-thread technology realizes this system..On procedure realization, we use the two-dimension array checkerboard form.This is mainly because of the consideration that the gobang's falling lotus and similar kinds of folk singing, will be not like Chinese chess such as move once more. At first the procedure can simply realize the judgment of loses and win. On this foundation, we improves it to be a programm which throughs the search to checkerboard's on node estimate value.According to each legitimate node's in estimate value, the computer may determine the final falling lotus and similar kinds of folk singing position, without thorough through calculation match falling lotus and similar kinds of folk singing position, thus decides own falling lotus and similar kinds of folk singing position. Through the optimized enhancement algorithm,the computer is able to decide its own falling lotus and similar kinds of folk singing position after guessing the possible falling lotus the competitor have.Thus enhanced the gobang game playing system's chess strength level.Systematic structure designs divide into three modules: The interfacial module, the estimate value module, searching module. In the concrete design, the estimate value module through more meticulous divideto complete.Different chessgame conditions have different estimation. These could appraise a relatively reasonable estimation which matches the chessgame condition.Gobang game playing system is implemented by JAVA language,and according to the understanding of SCOKET technology.By the multi-thread processing technology,we connect the server and the multi-client on the basis of it,it deals with the message deliver be the server and the client,and process the information between many clients using the multi-thread processing technology.The innovative research mainly has:1.In allusion to the problems of searching the depth endlessly,the limited spead and the low level of chess strength,even if theα-βsearch tree algorithm is working.We optimize and strengthen the algorithm and appraise threaten to revise estimate value, we also optimize using the genetic algorithm to the estimate value module.These cause the gobang game playing system, have the remarkable enhancement in the chess strength level and the aspect searching speed.2. The system designs the composition specially grade to fight the way , and different rank takes different algorithm, hence the optimized effect is obvious.The practice proved that compared with the classical one,the gobang game playing system realized in this passage have the remarkable enhancement in the chess strength level and the search speed aspect. At the same time, this design adopting the algorithm optimization method and the graduation to fight the way comparison optimization effect designing mentality, may bring the heuristic thought for other chess game algorithm researchers.The next , we mainly tocus on the memory searching aspect of the classical chessgame.Meanwhile,we want to realize the gobang international rules in the hand.Changing the chess and many other complicated regulation.Having done all of them,the system playing chess is more intellectualized and have higher efficency.