A global optimization by particle swarms with asymmetrical normal distribution

点击翻译按钮获取中文摘要

In recent years, the particle swarm optimization (PSO) has received much attention as a global optimization method. PSO succeeds in rapidly finding practically suboptimal solutions. However, it often fails to find the globally optimal solutions. It is therefore desirable to develop improved PSOs that rapidly find the globally optimal solutions with a high probability. As such an improved PSO, cautious particle swarm (CPS) has been proposed. The CPS employs an asymmetrical normal distribution to represent the probability density function for generating trajectories of the particles. In this research project, the effectiveness of the CPS and its contributing factors were examined.