火鹰优化算法Fire Hawk Optimizer (FHO) matlab代码

%% Fire Hawk Optimizer (FHO) source code - version 1.0 % Fire Hawk Optimizer: a novel metaheuristic algorithm % https://link.springer.com/article/10.1007/s10462-022-10173-w %% Clear Workspace & Command Window clc; clear all; %% Problem Information CostFunction = @(x) Sphere(x); % @ Cost Function VarNumber = 10; % Number of variables; VarMin = -10 *ones(1,VarNumber); % Lower bound of variable; VarMax = 10 *ones(1,VarNumber); % Upper bound of variable; %% General Parameters of the Algorithm MaxFes = 1500 ; % Maximum number of generations nPop = 25 ; % Maximum number of initial candidates HN = randi([1 ceil(nPop/5)],1,1) ; % Maximum number of FireHawks % Counters Iter=0; FEs=0; %% Initialization Pop=[]; Cost=[]; for i=1:nPop % Initialize Positions Pop(i,:)=unifrnd(VarMin,VarMax,[1 VarNumber]); % Cost Function Evaluations Cost(i,1)=CostFunction(Pop(i,:)); FEs=FEs+1; end % Sort Population [Cost, SortOrder]=sort(Cost); Pop=Pop(SortOrder,:); BestPop=Pop(1,:); SP=mean(Pop); % Fire Hawks FHPops=Pop(1:HN,:); % Prey Pop2=Pop(HN+1:end,:); % Distance between Fire Hawks and Prey for i=1:HN nPop2=size(Pop2,1); if nPop2<HN break end Dist=[]; for q=1:nPop2 Dist(q,1)=distance(FHPops(i,:), Pop2(q,:)); end [ ~, b]=sort(Dist); alfa=randi(nPop2); PopNew{i,:}=Pop2(b(1:alfa),:); Pop2(b(1:alfa),:)=[]; if isempty(Pop2)==1 break end end if isempty(Pop2)==0 PopNew{end,:}=[PopNew{end,:} ;Pop2]; end % Update Bests GB=Cost(1); BestPos=BestPop; %% Main Loop while FEs<MaxFes Iter=Iter+1; PopTot=[]; Cost=[]; for i=1:size(PopNew,1) PR=cell2mat(PopNew(i)); FHl=FHPops(i,:); SPl=mean(PR); Ir=unifrnd(0,1,1,2); FHnear=FHPops(randi(HN),:); FHl_new=FHl+(Ir(1)*GB-Ir(2)*FHnear); FHl_new = max(FHl_new,VarMin); FHl_new = min(FHl_new,VarMax); PopTot=[PopTot ;FHl_new]; for q=1:size(PR,1) Ir=unifrnd(0,1,1,2); PRq_new1=PR(q,:)+((Ir(1)*FHl-Ir(2)*SPl)); PRq_new1 = max(PRq_new1,VarMin); PRq_new1 = min(PRq_new1,VarMax); PopTot=[PopTot ;PRq_new1]; Ir=unifrnd(0,1,1,2); FHAlter=FHPops(randi(HN),:); PRq_new2=PR(q,:)+((Ir(1)*FHAlter-Ir(2)*SP)); PRq_new2 = max(PRq_new2,VarMin); PRq_new2 = min(PRq_new2,VarMax); PopTot=[PopTot ;PRq_new2]; end end for i=1:size(PopTot,1) Cost(i,1)=CostFunction(PopTot(i,:)); FEs=FEs+1; end % Sort Population [Cost, SortOrder]=sort(Cost); PopTot=PopTot(SortOrder,:); Pop=PopTot(1:nPop,:); HN = randi([1 ceil(nPop/5)],1,1); BestPop=Pop(1,:); SP=mean(Pop); FHPops=Pop(1:HN,:); Pop2=Pop(HN+1:end,:); clear PopNew % Distance between Fire Hawks and Prey for i=1:HN nPop2=size(Pop2,1); if nPop2<HN break end Dist=[]; for q=1:nPop2 Dist(q,1)=distance(FHPops(i,:), Pop2(q,:)); end [ ~, b]=sort(Dist); alfa=randi(nPop2); PopNew{i,:}=Pop2(b(1:alfa),:); Pop2(b(1:alfa),:)=[]; if isempty(Pop2)==1 break end end if isempty(Pop2)==0 PopNew{end,:}=[PopNew{end,:} ;Pop2]; end % Update Bests if Cost(1)<GB BestPos=BestPop; end GB=min(GB,Cost(1)); % Store Best Cost Ever Found BestCosts(Iter,1)=GB; % Show Iteration Information disp(['Iteration ' num2str(Iter) ': Best Cost = ' num2str(BestCosts(Iter))]); end Eval_Number=FEs; Conv_History=BestCosts; Best_Pos=BestPop; figure xlabel('迭代次数') ylabel('适应度值') plot( BestCosts)