摘要
为提高基本蝙蝠算法的局部最优解开发能力,拟引入速度权重扰动机制,提出一种基于速度权重扰动机制的改进蝙蝠优化算法.在算法迭代寻优过程中,蝙蝠个体自身当前位置优于群体当前位置均值时,选择带有速度权重扰动机制的速度演化策略更新下一代速度信息,从而提高算法跳出局部最优的能力,并最终实现群体逼近收敛到全局最优解.针对典型基准测试函数的仿真实验结果表明,该速度机制能够有效提高蝙蝠个体的局部开发能力,加强算法的全局寻优能力.
The velocity-weighted disturbance mechanism is intended to be introduced here and an improved bat algorithm is presented based on it to enhance the exploitation ability of local optimal solution with basic bat algorithm. In the iterative optimum searching process of the algorithm, the velocity information of next generation will be updated by the velocity evolution strategy with current position of bat individual proper is superior to the mean of current position of the population. Therefore, the ability of the algorithm to avoid of typical benchmark testing function local optimum will be improved and convergence to global optimum during it approximation will be realized. Simulation result shows that this velocity mechanism is effective in improving the local exploitation ability of the bat individual and enhancing the global optimum searching ability of the algorithm.
出处
《兰州理工大学学报》
CAS
北大核心
2016年第1期104-108,共5页
Journal of Lanzhou University of Technology
基金
国家自然科学基金(51365030)
甘肃省自然科学基金(2014G802715)
关键词
群集智能
蝙蝠算法
全局优化
权重扰动
swarm intelligence
bat algorithm
global optimization
weight disturbance