基于蜣螂优化的改进粒子群算法

针对标准粒子群算法存在的局部最优、早熟和慢收敛等问题,提出了一种新的粒子群更新方法。改进了算法惯性权重,引入一种新的更新方式;借鉴蜣螂优化算法中蜣螂滚球、繁殖、觅食和偷窃行为,将基本粒子群的操作划分为寻优、变异、波动和跳跃,从而提高了算法的全局寻优能力和收敛速度,并避免了早熟问题。通过与其他9种智能算法进行实验对比表明,在10个基准测试函数中,基于蜣螂优化的改进粒子群算法在寻优能力和收敛速度方面表现出色,证实了该算法的优越性。

The improved particle swarm optimization algorithm based on dung beetle optimization

To address the issues of local optimization, prematurity and slow convergence inherent in the standard particle swarm algorithm, we propose an improved particle swarm updating method. Firstly, the inertia weight is improved and a new updating method is introduced. Secondly, based on the behavior of rolling, breeding, foraging and stealing of Dung beetle optimization algorithm, the operation of basic particle swarm is divided into optimization, variation, fluctuation and jump, thus improving the global optimization ability and convergence speed of the algorithm, and avoiding the prematurity problem. Through experimental comparison with the other 9 intelligent algorithms, the results show that among the 10 benchmark test functions, the improved PSO based on Dung beetle optimization performs well in terms of optimization ability and convergence speed, thus confirming the superiority of this algorithm.