The hybrid algorithm for the study of geometric configurations of 2D clusters of uniformly charged classical particles

摘要

We present the modified, hybrid algorithm for the calculation of topological properties of 2D clusters of charged classical particles. The issue has been so far studied using simulation (Monte Carlo) as well as optimization (Genetic Algorithm) techniques. In the paper we combine the PSO global optimization mechanism with the local search algorithm. We propose different objective functions for different issues analyzed. We show the role of different phases of optimization. We are able to reproduce the earlier results and to present new configurations. It seems also that for the majority of issues our approach is fast and effective.