Hello Siamak,
According to my understanding you are working with “gamultiobj” function which is part of the Global Optimization Toolbox. In MATLAB, the “gamultiobj” function is used for multi-objective optimization, and while it offers several plotting functions, it does not directly provide a built-in plot function for convergence curves of individual objectives like “gaplotbestf” function. However, you can create custom plot functions to visualize the convergence of each objective.
To achieve this you can write a function that captures and plots the convergence of each objective over iterations. This function will be called at each generation and then integratethe custom plot function using the “options” parameter in “gamultiobj”. Here is an example:
function state = plotConvergence(options, state, flag)
numObjectives = size(state.Score, 2);
persistent bestScores;
if isempty(bestScores)
bestScores = inf(state.Generation, numObjectives);
end
for i = 1:numObjectives
bestScores(state.Generation + 1, i) = min(state.Score(:, i));
end
% Plot the convergence curve for each objective
if strcmp(flag, 'iter')
figure(1);
for i = 1:numObjectives
subplot(numObjectives, 1, i);
plot(bestScores(1:state.Generation + 1, i), 'LineWidth', 2);
title(['Convergence of Objective ', num2str(i)]);
xlabel('Generation');
ylabel(['Objective ', num2str(i), ' Value']);
grid on;
end
drawnow;
end
end
options = optimoptions('gamultiobj', ...
'PlotFcn', @plotConvergence, ...
'Display', 'iter');
% Define your objective function and constraints here
% objectiveFunction = @(x) [objective1(x), objective2(x)];
% lb = [lower_bounds];
% ub = [upper_bounds];
[x, fval] = gamultiobj(objectiveFunction, numVariables, [], [], [], [], lb, ub, options);
Hope this helps!
To know more about the “gamultiobj” function you can refer to the documentation link below:
