jMetal / jMetalPy

A framework for single/multi-objective optimization with metaheuristics
https://jmetal.github.io/jMetalPy/index.html
MIT License
497 stars 150 forks source link

How to avoid gabs in the solution #137

Closed flxzmmrmnn closed 5 months ago

flxzmmrmnn commented 2 years ago

I am using a SMPSO algorithm with the following parameters:

algorithm = SMPSO(
    problem=problem,
    swarm_size=200,
    mutation=PolynomialMutation(probability=1.0 / number_of_variables, distribution_index=20),
    leaders=CrowdingDistanceArchive(200),
    termination_criterion=StoppingByEvaluations(15000))

Everything is working so far and my results look as following:

Riyadh_MP_Opti_20220114_LCOE_all_Topologies

I am facing the problem that I have big gaps in my solution as can be seen in the graph above. In the green Pareto Front the distance between the solutions for Capacity Factors below 0.8 are much larger than for Capacity Factors above 0.8. The "green" problem has 6 variables and 2 objectives, the "blue" problem has only 3 variables. I am usind 15000 evaluations, a swarm size of 200 and a number of solutions of also 200.

How do I have to change my algorithm setting in order to get a more continuous Pareto Front? What is a good ratio between number of evaluations, swarm size, number of solutions and the number of variables?