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.
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.
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.
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.
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
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))]);
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 .
