A multi-objective particle swarm optimization algorithm based on dynamic boundary search for constrained optimization
Due to increased search complexity in multi-objective optimization, premature convergence becomes a problem. Complex engineering problems poses high number of variables with many constraints. Hence, more difficult benchmark problems must be utilized to validate new algorithms performance. A well-kno...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Published: |
Elsevier
2018
|
| Subjects: | |
| Online Access: | http://eprints.um.edu.my/21228/ https://doi.org/10.1016/j.asoc.2018.06.022 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Due to increased search complexity in multi-objective optimization, premature convergence becomes a problem. Complex engineering problems poses high number of variables with many constraints. Hence, more difficult benchmark problems must be utilized to validate new algorithms performance. A well-known optimizer, Multi-Objective Particle Swarm Optimizer (MOPSO), has a few weakness that needs to be addressed, specifically its convergence in high dimensional problems and its constraints handling capability. For these reasons, we propose a modified MOPSO (M-MOPSO) to improve upon these aspects. M-MOPSO is compared with four other algorithms namely, MOPSO, Multi-Objective Grey Wolf Optimizer (MOGWO), Multi-Objective Evolutionary Algorithm based on Decompositions (MOEA/D) and Multi-Objective Differential Evolution (MODE). M-MOPSO emerged as the best algorithm in eight out of the ten constrained benchmark problems. It also shows promising results in bioprocess application problems and tumor treatment problems. In overall, M-MOPSO was able to solve multi-objective problems with good convergence and is suitable to be used in real world problem. |
|---|