Diversity collaboratively guided random drift particle swarm optimization

The random drift particle swarm optimization (RDPSO) algorithm is an effective search technique inspired by the trajectory analysis of canonical PSO and free electron model in metal conductors placed external electric field. However, like other variants, RDPSO also inevitably encounters premature convergence when solving multimodal problems. To address this issue, paper proposes a novel diversity collaboratively guided (DCG) strategy for that enhances ability algorithm. In strategy, two kinds measures are defined modified collaborative manner. Specifically, whole process divided into three phases based on changes measures. each phase, different values selected key parameters update equation to make perform modes. Consequently, improved with DCG (DCG-RDPSO) can maintain its dynamically at certain level, thus constantly without stagnation until terminates. performance evaluation proposed done CEC-2013 benchmark suite, comparison several versions RDPSO, variants non-PSO evolutionary algorithms. Experimental results show significantly improve robustness most Further experiments economic dispatch problems verify effectiveness strategy.