Let us take an example of PSO implementation by Syed Ali Mirjalili, respected researcher in optimization field.
% Main loop
for t = 1 : maxIter
% Calcualte the objective value
for k = 1 : noP
currentX = Swarm.Particles(k).X;
position_history(k , t , : ) = currentX;
Swarm.Particles(k).O = fobj(currentX);
average_objective(t) = average_objective(t) + Swarm.Particles(k).O;
Here, why we use dot operator? Like this I face other issues.
In brief, what are your suggestions about optimization algorithm(s) implemetation.?
What ""dot"" are you talking about ?!
Do you mean this (Swarm.Particle) ?
In general, in programming this is called ""structure"" ..
You can group a number of properties for one structure, for example.. each particle has a vector represents the positions, and a value of velocity. Therefore, we can create a structure for the particles ..
Particle.Position
Particle.Velocity
General Form for the structure ( Structure . Property )
-------------------
currentX = Swarm.Particles(k).X;
this line means; store the value of the position (X) of the particle (k) in the swarm in a vector (currentX).
SO;
Swarm = Structure
Particles = An inside structure
X = A property represents the position of the particle.
You can read it like this: The position X of the particle (k) in the swarm (Swarm), store it in currentX
----------------------
This is another line:
Swarm.Particles(k).O = fobj(currentX);
Same as the previous explanation, calculate the fitness value using fobj for (currentX), and keep the result in the Objective value (O) for the particle (k) in the swarm (Swarm).
------------------------
In brief, what are your suggestions about optimization algorithm(s) implemetation.?
There are no specific suggestions.. you need to practice a lot, read a lot, in order to enhance/increase your programming skills, then you can implement any algorithm you need in your way. No need to do just like the others..
Hope this helps,
Regards..
Please see
https://www.youtube.com/channel/UCz6Q1WpIhQDD0Wvb7RlL9Sw/playlists
You need program them. Since metaheuristic maintain some similaraties. The framework can be the same.
Dear Sajad,
The ease of implementation relies on your programming skills & clear understanding of the method to be implemented. As Dr Sinan said you need to enhance your programming skills.
The following article would be good enough to this problem;
Research Particle Swarm Optimization: Algorithm and its Codes in MATLAB
Dear,
Please check the following links :
PARTICLE SWARM OPTIMIZATION (PSO) MATLAB CODE EXPLANATION
version 1.0.0.0 (1.85 KB) by Muhammad Raza
Minimize function using Particle Swarm Optimization
- Badges
- Science topic