Gaussian-Chaos-Enhanced Sub-dimensional Evolutionary Particle Swarm Optimization Algorithms for Global Optimization

In this paper, a sub-dimensional evolutionary of the particle swarm optimization algorithm based on Gaussian-chaos enhancement strategies (GC-SDPSO) is proposed. The advantage of the sub-dimensional evolutionary particle swarm algorithm is that each dimension of the particle is iterated in turn, rather than a whole. In order to overcome the shortcomings of sub-dimensional evolution algorithm, we introduce the complementary strategy of Gaussian and chaos in the iterative process. Gaussian mutation and chaos perturbation is performed with high aggregation degree to improve the global convergence and to ensure particle diversity. To evaluate the efficiency of the GC-SDPSO algorithm, we also choose nine different types of benchmark functions to test. Compared with SDPSO, GSDPSO, experimental results showed that GC-SDPSO is effective; convergence speed and the quality of the solution are improved.