Is proposed SMPSO algorithm an unconstrained multi-objective optimization?

Hi friends, First of all thanks for this great job and second unfortunately I have an issue with SMPSO algorithm … also I have to tell that I’m using .Net package (v4.5 -> that Obviously is not java based & also I know that it discontinued but I have to use it ! ) … Anyway I need to know that Is Speed-constrained Multi-objective PSO (known as SMPSO) an unconstrained multi-objective optimization? or not ?

P.S: I read the original paper (SMPSO: A New PSO Meta-heuristic for Multi-objective Optimization) and It’s used for unconstrained benchmark problems like ZD1-6 & DTLZ1-7 And there’s no point in using this algorithm in constrained issues … (unless I’m mistaken) Also I checked the c# code (jmetal\JMetalCSharp\Metaheuristics\SMPSO\SMPSO.cs) and there is a line for Problem.EvaluateConstraints(particle) :

//Evaluate the new particles in new positions
		for (int i = 0; i < particles.Size(); i++)
		{
			Solution particle = particles.Get(i);
			this.Problem.Evaluate(particle);
			this.Problem.EvaluateConstraints(particle);
		}

but indeed it seems there’s no limitation in adherence to restrictions because I used it for my constrained problem and Obviously algorithm doesn’t’ care about the constrain at all ! ( Feasibility Ratio approximately 16% for SMPSO compare to 70% and 80% for NSGA-II or SPEA2 !)

By the way , any hints and helps are acceptable 😃) thanks a lot

Sincerely Yours, Max

[ajnebro]

@3threeFX , sorry for not responding to your messages. I have not forgotten that, but this week is complicated to me and I have time to take a look at them.

[ajnebro]

In original paper (SMPSO: A New PSO Meta-heuristic for Multi-objective Optimization) the Chi is define with below equation :

You are right in that the equation we use is not same. As I mentioned before, we try a lot configurations and we ended up with the current one. We found out that when the particles reached out of the valid ranges of the variables, the best strategy was to reverse the direction.