A hybrid Particle Swarm Optimization considering accuracy and diversity of solutions

Particle Swarm Optimization (PSO) is an optimization method that emulates the behavior of creatures such as a flock of birds or a school of fish. Two typical PSO information exchange formats are the Gbest model and the Lbest model. The Gbest model is the most basic model, but this model can converge quickly on a solution and may become trapped at a local solution. On the other hand, the Lbest model converges slowly on the solution but its global search capability is better. In this study, we propose a method of remedying the drawback of PSO in that it tends to become trapped at a local solution, by maintaining the diversity of the search by a global search using the Lbest model in the early stages of the search, then switching to a local search by the Gbest model in the final stages. We also confirm the validity of this method by simulation experiments using benchmark problems. As a result, we confirmed that accuracy of discovery of the optimal solution was increased, although convergence on the solution was somewhat delayed.