Double Parameter Evaluation Function Based On Path

Computer game is a challenging research field that is full of vigor.It is an important research direction of artificial intelligence.It's the base of machine intelligence war gaming,intelligent decision system etc.Machine game is considered to be one of the most challenging research directions in the field of artificial intelligence.Computer game is often seen as a "fruit fly" of artificial intelligence.Its research results also have an important impact on other fields.Therefore,it has significant meaning on research of computer game.Connect6 is a new kind of chess which has been occurred in recent years.It has become more and more popular in recent years,attracting more and more researchers to study.Evaluation function is the core method for judging the situation of a position for the game program.The accuracy of evaluation function determines the accuracy of the procedure to judge the situation,so that the program can choose the right response plan.Generally,the evaluation methods can be roughly divided into two categories,based on the evaluation function of the chess type and the evaluation function based on the road.There is no good solution to the segmentation problem of connect6 when use evaluation function based on chess type.The segmentation of chess type is very complicated and costs too much time.The traditional road based evaluation function cannot deal with all kinds of situations well.In this paper,we study the evaluation function based on global path and the evaluation function based on local path.The combination of these two functions is an evaluation function that can accurately evaluate the overall situation and ensure the execution efficiency.At the same time,two groups of parameters are applied to our advantages and our disadvantages,in order to cope with different situations.In order to verify the rationality of double parameter evaluation function,this paper design the inclination experiment?the accuracy of estimation experiment and the power level experiment.The inclination experiment verifies that different experimental parameters will lead to different inclination.The estimation accuracy experiment verifies that in some situations,the double parameter evaluation function can more accurately judge the current situation than the traditional road based evaluation function,and can guide the game more accurately.The experiment of power level proves that using the double evaluation parameter evaluation function has a higher winning rate than the traditional road based evaluation function.In addition,due to the great influence of evaluation function on search efficiency,a search efficiency experiment is designed to compare the efficiency of two evaluation functions.At the end of this paper,based on the proposed double parameter evaluation function,we design and implement a connect6 program called “Executor”.The functions and implementation methods of each function module are introduced in detail,and the effectiveness of the method is further verified.