How are computer programs for chess and chess designed?

Chess defeated humans more than 10 years ago by Deep Blue, in recent years, the rating of chess software alone for personal computers has greatly surpassed that of humans, and there is no human world champion who is stupid enough to challenge himself with the computer. Originally, chess man-machine matches were often held, but since a few years ago, in order to make the game more suspenseful, the rules of the game have been changed to allow a pawn outside the f-line - if you have a more in-depth study of chess, you will know how valuable a pawn is. The structure of the chess software is divided into:

1. Human-computer interface: allow humans to directly talk to the computer in the language of chess;

2. Engine: Calculate the weight of the position score, and obtain the score value with a certain algorithm, positive numbers will give white an advantage, negative numbers will give black an advantage, and the larger the score (smaller), the better white (black) will be. The World Computer Chess Championship (WCCC) is held every year, and the computing power of the engine is measured in Moore's Law years, while the computing power of the human brain is measured in millions of years of evolution, and the gap will only widen, not to mention the efforts of programmers to optimize the algorithm;

3. Opening Library: Gathering the openings of millions of human chess games, so that under the premise of having an opening library, the computer can directly get out of the opening moves that humans think of without calculating. Chess grandmasters will remember quite a few openings, but computers will remember millions of games;

4. Endgame Library: The supercomputer calculates all the positions of the remaining pieces and stores them for calling. When the situation is stored in the remaining endgame library, there is no need to calculate, and the computer can know the result directly.

The endgame library within 6 subs has been calculated, and most of the 7 sub endgame libraries have been calculated. <>

The basic principles of chess software are decision trees and pruning algorithms, and the difference between different programs lies mainly in the details of these algorithms. I saw an answer before that it said enumeration, which is absolutely not possible. In a game of chess, there are about dozens or hundreds of possible moves for each move (of course, most of them are nonsensical moves, normal people will not think about it), the computing power of Deep Blue is 200 million moves per second, Kasparov can take into account the situation after 14 moves in the long exam, how long does it take to calculate Deep Blue if you only use the enumeration method?

If the numbers are too exaggerated, don't forget it. Therefore, it is necessary to prune the decision tree, that is, to reasonably ignore some unreasonable possibilities, there are many different pruning algorithms, such as - pruning, etc. (the link below is the application of the algorithm in Othello, better understanding), these are the application of the basic knowledge of computer science, there is a lot of information to consult, the description is more detailed than mine, so I will not say more. In conclusion, different pruning algorithms affect the speed and depth of the calculations.

The chess program is to let the intriguing play of chess with the computer, and I will introduce it to you below, welcome to read it.

Board games generally have a mathematical model, and all the computer does is determine each possible move for the next step of the move, as well as the next ...... of each moveAnd so on, that is, the computer can determine what to do next, so that it can benefit itself in the next step and even how many steps in the future. With the increase in the number of steps of the computer, the amount of calculation increases exponentially, and the calculation time will be longer, which is closed, in addition to the algorithm factors of chess programs, the test is the computer's CPU computing speed So it is necessary to take a compromise between the computer's "**ability" and the waiting time, such as 20 steps. In 1997, IBM's "Deep Blue" defeated the world chess champion Kasiparov because the computing power of his computer was the best in the world at that time.

Chinese chess source code - C language small programPage Change* unsigned char set svga mode***int vmode***Set svga screen mode* unsigned int get svga mode***Get current svga screen mode* drawbmp***int start x,int start y,char filename*fseek***fp,28,seek set***fread*** i,2,1,fp***if***i!=8***Check if it is a 256-color dot matrix * fseek***fp,18,seek set***fread***width,4,1,fp***fread***length,4,1,fp***if***firsttime***elsefseek***fp,300,seek set***k=**width%4***4-width%4***0; width correction value * for***j=length-1+start x; j>=start_x;j--*fread***buffer,width,1,fp***for***i=start_y,n=0;i

<> name of the game: Playing chess with a computer. Game Type: Casual puzzle.

Game size: 386k.

Interface language: Chinese game.

Game Description: Playing chess with the computer is a small chess and card game. Xiangqi is a chess art activity with a long history in China, with simple tools and strong fun.

It can be practiced in man-machine battles, two-player battles and classic chess games, and the archive reading function has been added, the interface is more friendly, and the operation is smoother.

It should be reminded that some games are multi-file, or there are **calls, or there are domain name access restrictions inside, and you will not be able to play normally after **, and you can only **play the game.

In the mid-'80s, Hans, a computer scientist at Carnegie-Mellon. Hans Berliner continued Ken Thompson's career. The scientist, who was the world champion of the Chess Communication Tournament, built a hardware-powered chess machine --- called "Hitech".

He and his student Carl Ebeling designed a hardware chess move generation chip. Equipped with 64 such chips, Hitech won the World Computer Chess Championship runner-up in 1986 by a narrow defeat to Cray.

Soon after, Berliner students Feng-Hsiung Hsu, Murray Campbell and others developed their own chess machine called --- "Chiptest" which later developed "Deep Thought", which was worth $5,000 and could count 500,000 positions per second. Xu Fengxiong and Campbell later broke away from their teachers and joined IBM. Together with Joe Hoane, he developed Deep Blue.

Deep Blue, which played against chess giants --- Kasparov in Philadelphia and New York, contained an IBM SP 2 server made up of a large number of specialized chips that could perform fast calculations. Each dedicated chip can handle between two million and three million situations per second. More than 200 of these chips are combined, and the programs running on them can process 200 million situations per second.