Scribd
Upload
45 views

TP 03: Technique D'optimisation PSO

The document describes implementing a particle swarm optimization (PSO) algorithm in MATLAB to optimize a sphere function. The PSO algorithm initializes a population of particles with random positions and velocities within bounds. It then iteratively updates each particle's velocity and position based on its personal best position and the global best position. The algorithm runs for a maximum number of iterations, tracking the best cost at each iteration, and returns the global best solution once complete.

Uploaded by

CEM Yelle N centre
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
45 views

TP 03: Technique D'optimisation PSO

The document describes implementing a particle swarm optimization (PSO) algorithm in MATLAB to optimize a sphere function. The PSO algorithm initializes a population of particles with random positions and velocities within bounds. It then iteratively updates each particle's velocity and position based on its personal best position and the global best position. The algorithm runs for a maximum number of iterations, tracking the best cost at each iteration, and returns the global best solution once complete.

Uploaded by

CEM Yelle N centre
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

UNIVERSITE Ahmed zabana RELIZANE

Institut des sciences et technologies Département de Génie Electrique


2 ème année Master : Electrotechnique Industrielle

TP 03 : technique d’optimisation PSO

1. Objectif du Tp :
L’objectif de ce TP est d’implémenter un tel système d’optimisation PSO sous matlab en
tirant du caractère matriciel de cet environnement.

2. Algorithme PSO
Particle Swarm Optimization Function File: Sphere(x)
function F1 = Sphere(x)
F1 = sum(x.^2);
end

%%Particle Swarm Optimization File Name Save as: pso.m


clear;
close all;
%% Fitness Function Calling
FitnessFunction=@(x) Sphere(x); % Fitness Function Calling

% Total Number of Decision Variables Used


nVar=10;

% Size of Decision Variables Matrix


VarSize=[1 nVar];

% Lower Bound
LowerBound =-10;

% Upper Bound
UpperBound = 10;

%% Parameters Initialization Phase

% Maximum Number of Iterations used.


MaxT=100;

% Total Number of Search Agents used.


PopulationSize = 10;

% Initialize PSO Parameters


% Inertia Weight
w=1;

% Inertia Weight Damping Ratio


wdamp=0.99;

1
UNIVERSITE Ahmed zabana RELIZANE
Institut des sciences et technologies Département de Génie Electrique
2 ème année Master : Electrotechnique Industrielle

% Personal Learning Coefficient


c1=1.5;

% Global Learning Coefficient


c2=2.0;

% Velocity Limits
VelMax=0.1*(UpperBound-LowerBound);
VelMin=-VelMax;

%% Initialization Position, Cost, Velocity, Best_Position, Best_Cost


empty_particle.Position=[];
empty_particle.Cost=[];
empty_particle.Velocity=[];
empty_particle.Best.Position=[];
empty_particle.Best.Cost=[];
particle=repmat(empty_particle,PopulationSize ,1);
GlobalBest.Cost=inf;

for i=1:PopulationSize
% Initialize Position for each search Agent in the search space
particle(i).Position=unifrnd(LowerBound,UpperBound,VarSize);
% Initialize Velocity for each search Agent in the search space
particle(i).Velocity=zeros(VarSize);
% Fitness Values Calculation for each search Agent in the search space
particle(i).Cost=FitnessFunction(particle(i).Position);
% Update Personal Best Position for the particles
particle(i).Best.Position=particle(i).Position;
particle(i).Best.Cost=particle(i).Cost;
% Update Global Best Position for each search Agent in the search space
if particle(i).Best.Cost<GlobalBest.Cost
GlobalBest=particle(i).Best;
end
end
BestCost=zeros(MaxT,1);
%% PSO Main Loop
for CurrentIteration=1:MaxT
for i=1:PopulationSize
% Update Velocity for each search Agent in the search space
particle(i).Velocity = w*particle(i).Velocity
+c1*rand(VarSize).*(particle(i).Best.Position-particle(i).Position)
+c2*rand(VarSize).*(GlobalBest.Position-particle(i).Position);
% Apply Velocity Limits
particle(i).Velocity = max(particle(i).Velocity,VelMin);
particle(i).Velocity = min(particle(i).Velocity,VelMax);
% Update Position for Each Particle
particle(i).Position = particle(i).Position + particle(i).Velocity;
% % Check Boundries [-10, 10]
Outside=(particle(i).Position<LowerBound |
particle(i).Position>UpperBound);
particle(i).Velocity(Outside)=-particle(i).Velocity(Outside);
particle(i).Position = max(particle(i).Position,LowerBound);
particle(i).Position = min(particle(i).Position,UpperBound);
% Fitness Values Calculation
particle(i).Cost = FitnessFunction(particle(i).Position);
% Update Personal Best

2
UNIVERSITE Ahmed zabana RELIZANE
Institut des sciences et technologies Département de Génie Electrique
2 ème année Master : Electrotechnique Industrielle

if particle(i).Cost<particle(i).Best.Cost
particle(i).Best.Position=particle(i).Position;
particle(i).Best.Cost=particle(i).Cost;
% Update Global Best
if particle(i).Best.Cost<GlobalBest.Cost
GlobalBest=particle(i).Best;
end
end
end

BestCost(CurrentIteration)=GlobalBest.Cost;
disp(['Current Iteration Number = ' num2str(CurrentIteration) ': Best Cost
Found = ' num2str(BestCost(CurrentIteration))]);

w=w*wdamp;

end
BestSol = GlobalBest;
%% Results
figure;
%plot(BestCost,'LineWidth',2);
semilogy(BestCost,'LineWidth',2);
xlabel('Iteration Numbers');
ylabel('Best Cost Found');
grid on;

3. Travail demandé
1. Exécuter ce programme
2. Expliquer les étapes de cet algorithme
3. Donner le critère d’arrêt
4. Donner le résultat final ainsi la courbe .

You might also like