数学实验七： 遗传算法 实验报告.doc

年级、专业 姓名 学号 名单序号 实验时间 MATLAB版本： 注：实验报告的最后一部分是实验小结与收获 实验七 遗传算法1 用Matlab编制另一个主程序Genetic2.m，求例1的在第二种终止条件下的最优解提示：一个可能的函数调用形式以及相应的结果为：Count,Result,BestMember=Genetic2(22,6,-x*x+2*x+0.5,-1,2,-2,0.01,0.00001)% 附录1 Genetic2.mfunction Count,Result,BestMember=Genetic2(MumberLength,MemberNumber,FunctionFitness,MinX,MaxX,Fmin,MutationProbability,Precision) Population=PopulationInitialize(MumberLength,MemberNumber); Error=Precision+1; global Count; global CurrentBest; Count=1; PopulationCode=Population; PopulationFitness=Fitness(PopulationCode,FunctionFitness,MinX,MaxX,MumberLength); %用于计算群体中每一个染色体的目标函数值PopulationFitnessF=FitnessF(PopulationFitness,Fmin); %用于计算每个染色体的适应函数值PopulationProbability=Probability(PopulationFitnessF); %用于计算群体中每个染色体的入选概率Population,CurrentBest,EachGenMaxFitness=Elitist(PopulationCode,PopulationFitness,MumberLength); %用到最佳个体保存方法（“优胜劣汰”思想）EachMaxFitness(Count)=EachGenMaxFitness; MaxFitness(Count)=CurrentBest(length(CurrentBest); while ErrorPrecision NewPopulation=Select(Population,PopulationProbability,MemberNumber); Population=NewPopulation; NewPopulation=Crossing(Population,FunctionFitness,MinX,MaxX,MumberLength); Population=NewPopulation; NewPopulation=Mutation(Population,MutationProbability); Population=NewPopulation; PopulationFitness=Fitness(Population,FunctionFitness,MinX,MaxX,MumberLength); PopulationFitnessF=FitnessF(PopulationFitness,Fmin); PopulationProbability=Probability(PopulationFitnessF); Count=Count+1; NewPopulation,CurrentBest,EachGenMaxFitness=Elitist(Population,PopulationFitness,MumberLength); EachMaxFitness(Count)=EachGenMaxFitness; MaxFitness(Count)=CurrentBest(length(CurrentBest); Error=sum(abs(PopulationProbability-mean(PopulationProbability); Population=NewPopulation; end Dim=size(Population); Result=ones(2,Dim(1); for i=1:Dim(1) Result(1,i)=Translate(Population(i,:),MinX,MaxX,MumberLength); end Result(2,:)=PopulationFitness; BestMember(1,1)=Translate(CurrentBest(1:MumberLength),MinX,MaxX,MumberLength); BestMember(2,1)=CurrentBest(MumberLength+1); close all subplot(211) plot(EachMaxFitness) subplot(212) plot(MaxFitness) Count,Result,BestMember=Genetic2(22,6,-x*x+2*x+0.5,-1,2,-2,0.01,0.00001)Count = 11Result = 0.9748 0.9748 0.9748 0.9748 0.9748 0.9748 1.4994 1.4994 1.4994 1.4994 1.4994 1.4994BestMember = 0.97481.49942 按照例2的具体要求，用遗传算法求上述例2的最优解 Count,Result,BestMember=Genetic1(20,6,-x*x+x+3,0,2,1,0.01,50)Count = 50Result = 0.6880 0.6880 0.6880 0.6880 0.6880 0.6880 3.2146 3.2146 3.2146 3.2146 3.2146 3.2146BestMember = 0.6880 3.21463 附录9子程序 Crossing.m中的第8行至第13行的程序表明，当Dim(1)=3时，将交换数组Population的最后两行，即交换最后面的两个个体其目的是什么？4 设，求 ，要设定求解精度到15位小数 Count,Result,BestMember=Genetic2(22,6,-x*x-4*x+1,-2,2,-12,0.01,1e-15)Count = 15Result =