A New Quantum Soliton Inspired Swarm Optimization Algorithm for Multi-objective Optimization

This paper introduces a novel approach to extending the multi-objective particle swarm optimization algorithm (MOPSO) characterized by the quantum concept of particle-like solitons, which are the solutions of the quantum nonlinear Schrödinger equation. Considering the motion scenario of the present algorithm based on the corresponding probability density function of quantum solitons allows producing new particle positions that overcome the deficiency of particles readily gathering in identical solutions and escaping from local Parato front. We examine the proposed algorithm over a set of known benchmark functions to evaluate the efficiency. Moreover, to achieve a more comprehensive conclusion about the performance, we compare it with the results obtained by two different state-of-the-art multi-objective metaheuristics, namely multi-objective particle swarm optimization (MOPSO) and non-dominated sorting genetic algorithm (NSGA-II). The numerical experiments show that in terms of accuracy, convergence, diversity, and distribution, the proposed algorithm provides promising results compared with other existing multi-objective optimization algorithms.