【免费分享】2024最新优化算法-花斑翠鸟优化算法PKO

斑翠鸟是一种美丽的鸟类，隶属于翠鸟科，主要分布在东南亚地区，如马来西亚、泰国、印度尼西亚等国。斑翠鸟的身体呈蓝色，头部有黑色斑点，翅膀和尾巴同样为蓝色。它们具有长而尖的嘴巴，适合捕食小型水生动物。斑翠鸟偏爱栖息在水边，常见于河流、湖泊和沼泽附近，以鱼类为主要食物，通常通过潜水来捕获猎物。捕食时，斑翠鸟会从栖息地的树枝上俯冲而下，迅速抓住猎物后返回树枝进食。

斑翠鸟优化算法（Pied Kingfisher Optimizer, PKO）是一种由Abdelazim Hussien于2024年提出的新型群体基元启发式算法。该算法受到斑翠鸟独特的狩猎行为和共生关系的启发，围绕三个阶段展开：栖息/悬停猎物（探索/多样化）、潜水寻找猎物（开发/集约化）以及培养共生关系。这些行为被转化为数学模型，用于有效解决不同搜索空间中的多种优化问题。

代码如下

function [Best_fitness,Best_position,Convergence_curve]=PKO(Popsize,Maxiteration,LB,UB,Dim,Fobj)
tic;
BF=8;%beating Factor
Crest_angles=2*pi*rand;
X=Initialization(Popsize,Dim,UB,LB);
Fitness = zeros(1,Popsize);
Convergence_curve=zeros(1,Maxiteration);
for i=1:Popsize
    Fitness(i)=Fobj(X(i,:));
end
% Calculate the fitness values of initial PKO.
[~,sorted_indexes]=sort(Fitness);
Best_position=X(sorted_indexes(1),:);
Best_fitness = Fitness(sorted_indexes(1));
Convergence_curve(1)=Best_fitness;
t=1;
PEmax=0.5;
PEmin=0;
while t<Maxiteration+1
    o=exp(-t/Maxiteration)^2;
    for i=1:Popsize
          if rand<.8  %exploration
              j=i;
              while i==j
              seed=randperm(Popsize);
               j=seed(1);
              end
               beatingRate=rand*(Fitness(j))/(Fitness(i));
              alpha=2*randn(1,Dim)-1;

              if rand<.5
                 T=beatingRate-((t)^(1/BF)/(Maxiteration)^(1/BF));
                 X_1(i,:)=X(i,:)+alpha.*T.*(X(j,:)-X(i,:));
              else
                  T= (exp(1)-exp(((t-1)/Maxiteration)^(1/BF)))*(cos(Crest_angles));
                  X_1(i,:)=X(i,:)+alpha.*T.*(X(j,:)-X(i,:));
            
              end
          else %exploitation
               alpha=2*randn(1,Dim)-1;
               b=X(i,:)+o^2*randn.*Best_position; 
               HuntingAbility=rand*(Fitness(i))/(Best_fitness);
               X_1(i,:)=X(i,:)+ HuntingAbility*o*alpha.*(b-Best_position);
      
          end
    end
      for i=1:Popsize

        FU=X_1(i,:)>UB;FL=X_1(i,:)<LB;X_1(i,:)=(X_1(i,:).*(~(FU+FL)))+UB.*FU+LB.*FL;
        fitnessn(i) = Fobj(X_1(i,:));
        if fitnessn(i)<Fitness(i)
            Fitness(i)=fitnessn(i);
            X(i,:)=X_1(i,:);
        end
        if Fitness(i)<Best_fitness
            Best_fitness=Fitness(i);
            Best_position=X(i,:);
        end
      end
      
      % Commensal association of the pied kingfisher with foraging Eurasian otters
      PE=PEmax-(PEmax-PEmin)*(t/Maxiteration);   %Predatory Efficiency
      for i=1:Popsize
           alpha=2*randn(1,Dim)-1;
         if rand>(1-PE)  
            X_1(i,:)=X(randi([1,Popsize]),:)+o*alpha.*abs(X(i,:)-X(randi([1,Popsize]),:));
         else
            X_1(i,:)=X(i,:);
         end
         FU=X_1(i,:)>UB;FL=X_1(i,:)<LB;X_1(i,:)=(X_1(i,:).*(~(FU+FL)))+UB.*FU+LB.*FL;
         fitnessn(i) = Fobj(X_1(i,:));
        if fitnessn(i)<Fitness(i)
            Fitness(i)=fitnessn(i);
            X(i,:)=X_1(i,:);
        end
        if Fitness(i)<Best_fitness
            Best_fitness=Fitness(i);
            Best_position=X(i,:);
        end
      end
   
    Convergence_curve(t)=Best_fitness;
    t = t + 1;
end
time = toc;
end