Chess game AI algorithm of chess and board source building tutorial

Chess game client implementation using Flash 9 development, the service side uses WIN32+VC6 development (based on IOCP), the database gateway using WIN32+VC6 development (based on IOCP,MYSQL5 implementation of the processing thread pool and database connection pool). Although the server was completed last year, But the corresponding robot AI algorithm has not been able to achieve. Get it on the blog today and hope to have the opportunity to explore with interested brothers.

Kevin in his blog gave his realization, the algorithm thought is to use the width of priority to generate a search tree, and then according to the skill of playing cards pruning and judgment right, the robot AI can be like the formation of the game, the realization of slowly growing under the ravages, Level gradually improve. Although Kevin gives the basic idea of the algorithm, but the younger brother is still not able to understand Zen Ah (more detailed algorithm chess source set up www.yasewl.com).

I think the game AI should have the following features:

(1) Team awareness

Playing Mahjong game each player playing independently, but for a game like the landlord, players are grouped, Ai robot and when the farmer's card strategy should be different.

(2) Note card

Ai Bots are able to respond to cards already played by individual players for future decision-making. At the same time, you should also provide an interface for server cheating. For example, Mahjong games, AI robots in the decision-making can secretly look at other players in the hands of the cards or to see the cards on the table has not been distributed; for the game of landlords, Ai bots can also secretly look at the hands of other players when making decisions.

(3) Training and learning

Through the practice with the game Master can train and improve the intelligence level of AI robot, gradually learn and improve.

(4) Storage of Knowledge Base

The knowledge base of an AI robot should be stored in the form of a data file in a computer. The robot loads the configured repository data file at boot time.

I do not know whether my idea is too ideal, anyway, still haven't found a suitable model to abstract and represent the game decision-making process

This article is from the "daytime do not know Night Black" blog, please be sure to keep this source http://13134261.blog.51cto.com/13124261/1947876

Chess game AI algorithm of chess and board source building tutorial