Abstract
A multi-objective particle swarm optimization (MOPSO) is known as a multipoint-search-based meta-heuristic approach to find diverse Pareto solutions efficiently, but has difficulty to handle the constraint conditions. Overcoming the disadvantage, this study proposes a hybrid algorithm incorporating MOPSO and sensitivity analysis on constrained conditions. When the design candidate is violated, the design candidate is moved to feasible domain based on the gradient information of the constraint conditions. The transerring algorithm is adopted for each design candidate at each iteration in MOPSO algorithms. Through several numerical examples, the diversity and convergency of the Pareto solutions and the performance of the proposed method are investigated.
