基于Mamdani模糊规则推理的自适应遗传算法matlab仿真

clc; clear; close all; warning off; addpath(genpath(pwd)); fismat = readfis('d.fis'); Npop = 100; MaxPop = 300; Ngeneration = 200; Pc = 0.7; Pm = 0.9; CrossPercent = 50; MutatPercent = 10; CrossNum = round(CrossPercent/100*Npop); if mod(CrossNum,2)~=0; CrossNum = CrossNum - 1; end MutatNum = round(MutatPercent/100*Npop); ElitNum = Npop - CrossNum - MutatNum; VarMin = 0; VarMax = 10; Nvar = 2; CostFuncName = @Cost_Fcn; Pop = rand(Npop,Nvar) * (VarMax - VarMin) + VarMin; Cost = feval(CostFuncName,Pop); [Cost Indx] = sort(Cost,'descend'); Pop = Pop(Indx,:); %Initialize prog parameter BestC = zeros(Ngeneration,1); MeanC = 0*BestC; for Iter = 1:Ngeneration % Elitism ElitPop = Pop(1:ElitNum,:); % Cross Over CrossPop = zeros(CrossNum,Nvar); k = 1; ParentIndexes = SelectParents_Fcn(Cost,CrossNum); rand_cross=rand(floor(CrossNum/2),1); for ii = 1:CrossNum/2 if rand_cross(ii) > Pc Par1Indx = ParentIndexes(ii*2-1); Par2Indx = ParentIndexes(ii*2); Par1 = Pop(Par1Indx,:); Par2 = Pop(Par2Indx,:); Son = CrossOver_Fcn(Par1,Par2); CrossPop(k:k+1,:) = Son; k = k+2; end end CrossPop = CrossPop(1:k-1,:); % Mutation k = 1; indx = randperm(Npop); for ii = 1:Npop if rand > Pm C = Pop(indx(ii),:); MutatPop(k:k+1,:) = Mutation_Fcn(C,VarMax,VarMin,Iter,Ngeneration); k = k+1; end if k > MutatNum,break,end end MutatPop = MutatPop(1:k-1,:); % New Population Pop = [ElitPop ; CrossPop ; MutatPop;Pop(ElitNum+1:end,:)]; l = size(Pop,1); if l > MaxPop Npop = MaxPop; Pop = Pop(1:Npop,:); else Npop = l; end Cost = feval(CostFuncName,Pop); [Cost Indx] = sort(Cost,'descend'); Pop = Pop(Indx,:); % Algorithm Progress varcost = mean(var(Pop)); dP = evalfis(varcost,fismat); Pm = Pm + dP(1); Pc = Pc + dP(2); if Pc > .9, Pc = .9; elseif Pc < .1, Pc = .1;end if Pm > .97, Pm = .97; elseif Pm < .03, Pm = .03;end % Compute number of CrossOver & Mutation MutatNum = round(MutatPercent/100*Npop); CrossNum = round(CrossPercent/100*Npop); BestC(Iter) = Cost(1); MeanC(Iter) = mean(Cost); % Show Result plot(BestC(1:Iter),'r','linewidth',2); hold on plot(MeanC(1:Iter),'k','linewidth',2); hold off figure(gcf) end figure plot(BestC/max(BestC),'r','linewidth',2); hold on plot(MeanC/max(BestC),'k','linewidth',2); hold off figure(gcf) legend('Best Individual','Mean of Individuals') xlabel('Generations') ylabel('Normilized Fitness') ylim([-.2,1.2]) %% Results disp('----------------------------------------------') BestSolution = Pop(1,:); BestCost = Cost(1,:); disp(['BestSolution = ',num2str(BestSolution)]) disp(['BestCost = ',num2str(BestCost)]) disp('----------------------------------------------') %% End of Program