查询结果:   易正俊,韦磊鹏,袁玉兴.自适应重生鱼群优化算法[J].计算机应用与软件,2016,33(6):227 - 230,276.
中文标题
自适应重生鱼群优化算法
发表栏目
算法
摘要点击数
737
英文标题
ADAPTIVE REBORN FISH SCHOOL OPTIMISATION ALGORITHM
作 者
易正俊 韦磊鹏 袁玉兴 Yi Zhengjun Wei Leipeng Yuan Yuxing
作者单位
重庆大学数学与统计学院 重庆 401331 重庆科技学院数理学院 重庆 401331    
英文单位
College of Mathematics and Statistics,Chongqing University,Chongqing 401331,China School of Mathematics and Science,Chongqing University of Science and Technology,Chongqing 401331,China    
关键词
人工鱼群算法 鱼群重生 正态分布 动态拥挤度因子 优化
Keywords
Artificial fish school algorithm Fish rebirth Normal distribution Dynamic crowding factor Optimisation
基金项目
国家自然科学基金项目(11371384,6967 4012);重庆市科技攻关计划项目(CSTC2009AC3037)
作者资料
易正俊,教授,主研领域:人工智能,智能算法,信息融合与处理。韦磊鹏,硕士生。袁玉兴,助教。 。
文章摘要
针对传统人工鱼群算法求解高维优化问题收敛速度较慢,易于陷入局部最优,提出自适应重生鱼群优化算法。首先在每次迭代过程中,不断地给鱼群注入“新生命”使鱼群得以重生;然后采用正态分布动态调整拥挤度因子的上限值使得算法更贴近于鱼群搜索食物的过程。实验结果表明,改进后的算法既保证收敛速度、增加算法获得全局最优的可能性,又适用于求解大规模的优化问题。其中的两个算例采用改进的鱼群算法进行优化,优化结果与实际具有良好的一致性,说明了改进算法的有效性和实用性。
Abstract
Traditional artificial fish school algorithm converges slowly and is prone to falling into local optimum when solving high-dimensional optimisation problems. In light of this, we presented the adaptive reborn fish school optimisation algorithm. First, in the process of each iteration we injected the "new life" into fish school incessantly, which made the rebirth of the fish school; Then we used normal distribution to dynamically adjust the upper threshold of crowding factor, making the algorithm more close to the process of fish school’s forage. Experimental result showed that the improved algorithm ensured the convergence speed and increased the probability of the algorithm in obtaining global optimum, yet it were also suitable for solving large-scale optimisation problems. Two examples in the paper were optimised by the improved fish school algorithm, the optimisation results were in good conformity with the reality, this illustrated the effectiveness and practicability of the improved algorithm.
下载PDF全文