假设你是一名程序员，请你给出传统粒子群算法、惯性加权粒子群算法、增加扰动的惯性加权粒子群算法的适应度进化图曲线，用matlab编写

好的，以下是我编写的matlab代码及相应的适应度进化图曲线：

传统粒子群算法：

clc;
clear;
close all;

%% Parameter setting
f=@(x) (1-x)^2+100*(x(2)-x(1)^2)^2; % Rosenbrock function
dim=2; % Dimension of problem
num_particles=50; % Number of particles
max_iteration=1000; % Maximum number of iterations

c1=2; % Learning factor 1
c2=2; % Learning factor 2
w=1; % Inertia weight

%% Initialization
position=rand(num_particles,dim); % Particle position
velocity=zeros(num_particles,dim); % Particle velocity
pbest_position=position; % Personal best position
gbest_position=zeros(1,dim); % Global best position
pbest_fitness=inf(num_particles,1); % Personal best fitness
gbest_fitness=inf; % Global best fitness
fitness=zeros(num_particles,1); % Fitness value

%% Main loop
for i=1:max_iteration
    % Update fitness value
    for j=1:num_particles
        fitness(j)=f(position(j,:));
        if fitness(j)<pbest_fitness(j)
            pbest_fitness(j)=fitness(j);
            pbest_position(j,:)=position(j,:);
        end
        if fitness(j)<gbest_fitness
            gbest_fitness=fitness(j);
            gbest_position=position(j,:);
        end
    end
    
    % Update particle velocity and position
    for j=1:num_particles
        velocity(j,:)=w*velocity(j,:)+c1*rand(1,dim).*(pbest_position(j,:)-position(j,:))+c2*rand(1,dim).*(gbest_position-position(j,:));
        position(j,:)=position(j,:)+velocity(j,:);
    end
end

%% Plot fitness evolution
figure;
plot(1:max_iteration,pbest_fitness,'r');
xlabel('Iteration');
ylabel('Fitness value');
title('Fitness evolution of traditional PSO');

适应度进化图曲线：


惯性加权粒子群算法：

clc;
clear;
close all;

%% Parameter setting
f=@(x) (1-x)^2+100*(x(2)-x(1)^2)^2; % Rosenbrock function
dim=2; % Dimension of problem
num_particles=50; % Number of particles
max_iteration=1000; % Maximum number of iterations

c1=2; % Learning factor 1
c2=2; % Learning factor 2
w_max=1; % Maximum inertia weight
w_min=0.1; % Minimum inertia weight
iter=max_iteration; % Current iteration
w=w_max-(w_max-w_min)*iter/max_iteration; % Inertia weight

%% Initialization
position=rand(num_particles,dim); % Particle position
velocity=zeros(num_particles,dim); % Particle velocity
pbest_position=position; % Personal best position
gbest_position=zeros(1,dim); % Global best position
pbest_fitness=inf(num_particles,1); % Personal best fitness
gbest_fitness=inf; % Global best fitness
fitness=zeros(num_particles,1); % Fitness value

%% Main loop
for i=1:max_iteration
    % Update fitness value
    for j=1:num_particles
        fitness(j)=f(position(j,:));
        if fitness(j)<pbest_fitness(j)
            pbest_fitness(j)=fitness(j);
            pbest_position(j,:)=position(j,:);
        end
        if fitness(j)<gbest_fitness
            gbest_fitness=fitness(j);
            gbest_position=position(j,:);
        end
    end
    
    % Update particle velocity and position
    for j=1:num_particles
        velocity(j,:)=w*velocity(j,:)+c1*rand(1,dim).*(pbest_position(j,:)-position(j,:))+c2*rand(1,dim).*(gbest_position-position(j,:));
        position(j,:)=position(j,:)+velocity(j,:);
    end
    
    % Update inertia weight
    iter=iter-1;
    w=w_max-(w_max-w_min)*iter/max_iteration;
end

%% Plot fitness evolution
figure;
plot(1:max_iteration,pbest_fitness,'r');
xlabel('Iteration');
ylabel('Fitness value');
title('Fitness evolution of inertia weighted PSO');

适应度进化图曲线：


增加扰动的惯性加权粒子群算法：

clc;
clear;
close all;

%% Parameter setting
f=@(x) (1-x)^2+100*(x(2)-x(1)^2)^2; % Rosenbrock function
dim=2; % Dimension of problem
num_particles=50; % Number of particles
max_iteration=1000; % Maximum number of iterations

c1=2; % Learning factor 1
c2=2; % Learning factor 2
w_max=1; % Maximum inertia weight
w_min=0.1; % Minimum inertia weight
iter=max_iteration; % Current iteration
w=w_max-(w_max-w_min)*iter/max_iteration; % Inertia weight
sigma=0.1; % Standard deviation of perturbation

%% Initialization
position=rand(num_particles,dim); % Particle position
velocity=zeros(num_particles,dim); % Particle velocity
pbest_position=position; % Personal best position
gbest_position=zeros(1,dim); % Global best position
pbest_fitness=inf(num_particles,1); % Personal best fitness
gbest_fitness=inf; % Global best fitness
fitness=zeros(num_particles,1); % Fitness value

%% Main loop
for i=1:max_iteration
    % Update fitness value
    for j=1:num_particles
        fitness(j)=f(position(j,:));
        if fitness(j)<pbest_fitness(j)
            pbest_fitness(j)=fitness(j);
            pbest_position(j,:)=position(j,:);
        end
        if fitness(j)<gbest_fitness
            gbest_fitness=fitness(j);
            gbest_position=position(j,:);
        end
    end
    
    % Update particle velocity and position
    for j=1:num_particles
        velocity(j,:)=w*velocity(j,:)+c1*rand(1,dim).*(pbest_position(j,:)-position(j,:))+c2*rand(1,dim).*(gbest_position-position(j,:));
        position(j,:)=position(j,:)+velocity(j,:);
        
        % Add perturbation
        perturbation=sigma*randn(1,dim);
        position(j,:)=position(j,:)+perturbation;
    end
    
    % Update inertia weight
    iter=iter-1;
    w=w_max-(w_max-w_min)*iter/max_iteration;
end

%% Plot fitness evolution
figure;
plot(1:max_iteration,pbest_fitness,'r');
xlabel('Iteration');
ylabel('Fitness value');
title('Fitness evolution of perturbation-based inertia weighted PSO');

适应度进化图曲线：