一种基于种群多样性的自适应粒子群算法

An Adaptive Particle Swarm Optimization Algorithm Based on the Measurements of Population Diversity

以信息熵的角度研究了种群多样性测度的指标,提出了一种新的自适应粒子群算法.通过对种群多样性测度新指标的应用,采用保留最优个体的精英保留变异操作、新的速度项和动态惯性权重等技术,有效提高了种群的多样性.仿真试验说明了本文算法的优点.

From view of information entropy, this paper presents a new evolutionary algorithm-adaptive particle swarm optimization algorithm , which is based on the measurements of population diversity. Using the method of applying new measurements of particle swarm diversity, the algorithm uses a special mutation operator, and uses the elite individual reserved strategy, new velocity term and dynamic inertia weight to increase the swarm population diversity. The benchmark tests have showed the power of the algorithm in the numerical experiments.