Research And Design Of Genetic Algorithm And Simulated Annealing Algorithm In Scissor Stone Cloth Game

The game rock-paper-scissors as a kind of finger-guessing game is supposed to produce random results for decision making. But sometimes it is not random, because players can make decision according to their previous experience. We implemented the game into a computer program and gave the program artificial intelligence according to the "experience" of computer to simulate human’s strategy. In the present study the game of rock-paper-scissors is simulated by genetic algorithm and simulated annealing algorithm, and these two algorithms are implemented in the Visual Studio compiling environment using C/C++programming language. By comparing the different parameters and win rates between genetic algorithm and simulated annealing algorithm the better algorithm and parameters are selected. The main work of the following aspects:First, the game of rock-paper-scissors is modeled to meet the requirement for implementation of algorithms; Second, simple random strategy, imitating rival game strategy, and two intelligent algorithms were compared under experimental environment, the results showed that the two algorithms on the game model has a higher winning percentage than other strategies;Third, further research and analysis were applied to the two algorithms. The results show that:under the same conditions, the use of genetic algorithms game winning percentage is higher than the simulated annealing algorithm;Fourth, in order to increase the performance of the simulated annealing algorithm, a fusion algorithm were proposed and implemented, the results show that the fusion algorithm has a good effect to enhance the game’s winning percentage;Fifth, in order to solve the efficiency problem of the fusion algorithm, the relevant parameters are changed. The experimental results showed that:the efficiency of the fusion algorithm improved while the winning percentage was remained.