李顺, 郭星. 一种改进的自适应步长的萤火虫算法[J]. 微电子学与计算机, 2018, 35(8): 93-96, 100.
引用本文: 李顺, 郭星. 一种改进的自适应步长的萤火虫算法[J]. 微电子学与计算机, 2018, 35(8): 93-96, 100.
LI Shun, GUO Xing. An Improved Glowworm Swarm Optimization Alogorithm with Adaptive Step[J]. Microelectronics & Computer, 2018, 35(8): 93-96, 100.
Citation: LI Shun, GUO Xing. An Improved Glowworm Swarm Optimization Alogorithm with Adaptive Step[J]. Microelectronics & Computer, 2018, 35(8): 93-96, 100.

一种改进的自适应步长的萤火虫算法

An Improved Glowworm Swarm Optimization Alogorithm with Adaptive Step

  • 摘要: 提出一种预防萤火虫失活的自适应步长的萤火虫算法(PA-GSO).第一, 步长采用非线性递减方式, 初期步长较大移动速度快, 后期步长逐渐减少至固定的较小值, 实现了算法精度和速度的平衡.第二, 为了应对萤火虫的邻居集合为空集时可能丧失移动能力的现象, 采用了预防萤火虫失活机制优化萤火虫运动.通过实验对比GSO, A-GSO和CSGSO算法, 各方面指标验证了PA-GSO算法在寻优精度、收敛速度和稳定性等方面的提升.

     

    Abstract: An Preventing-inactivation adaptive-step GSO (PA-GSO) is proposed to prevent firefly inactivation. The first step is non-linear decreasing method.The initial step is larger and the moving speed is faster, and the later step is gradually reduced to a fixed smaller value, which realizes the balance of precision and speed in algorithm. And then in order to cope with the firefly's neighbor with the empty set may lose the mobility, using the prevention of firefly inactivation mechanism to optimize the firefly movement. By comparing the GSO, A-GSO and CSGSO algorithms, the PA-GSO algorithm is further improved in optimization accuracy, convergence speed and stability.

     

/

返回文章
返回