Particle swarm optimization algorithm with adaptive two-population strategy

The particle swarm optimization (PSO) algorithm is a intelligence (SI) used to solve problems. Owing its advantages in simplicity, using only few parameters, PSO has become one of the most popular algorithms. However, single structure leads challenges finding appropriate optima, resulting low convergence accuracy. To defects PSO, it necessary increase diversity populations involved as well enhance ability develop locally. In this study, we propose with an adaptive two-population strategy (PSO-ATPS), which adaptively divides population into two groups representing excellent and ordinary populations. Inspired by animal hunting behavior, new velocity–position update method proposed for general population. A velocity formulation decreasing inertia weights based on logistic chaotic mapping applied increases continuously changing search particles. addition, neighborhood (oscillation strategy) proposed, searches randomly own when motion stagnant updates position elite strategy. Among several strategies, PSO-ATPS achieved first place 7, 8, 9 tests involving 10 test functions 3 dimensions, indicating accuracy effectiveness PSO-ATPS. results show that performance competitive, many improvements developed can be PSO-ATPS, demonstrating potential further development.