A New Particle Swarm Optimization Method Enhanced With a Periodic Mutation Strategy and Neural Networks

Abstract

Particle swarm optimization (PSO), a relatively new population-based intelligence algorithm, exhibits good performance on optimization problems. However, during the optimization process, the particles become more and more similar, and gather into the neighborhood of the best particle in the swarm, which makes the swarm prematurely converged most likely around the local solution. A new optimization algorithm called multifrequency vibrational PSO is significantly improved and tested for two different test cases: optimization of six different benchmark test functions and direct shape optimization of an airfoil in transonic flow. The algorithm emphasizes a new mutation application strategy and diversity variety, such as global random diversity and local controlled diversity. The results offer insight into how the mutation operator affects the nature of the diversity and objective function value. The local controlled diversity is based on an artificial neural network. As far as both the demonstration cases' problems are considered, remarkable reductions in the computational times have been accomplished.