摘要
在求解多峰复杂函数的过程中,传统的模拟退火算法和禁忌搜索算法经常出现算法快速收敛于局部最优解、后期收敛速度变慢和搜索能力变差等问题.为解决这些问题,本文给出函数复杂度的定义,并提出基于函数复杂度的自适应模拟退火和禁忌搜索算法.该算法首先根据函数复杂度自适应调整步长控制参数,然后根据调整后步长求得函数的粗糙解,在此基础上再使用初始步长求得全局最优解.实验表明,该算法不仅可以跳出局部最优解的限制,并且减少了迭代次数,有效地提高了全局和局部搜索能力.
In the process of applying traditional simulated annealing algorithm and tabu search algorithm to solving multi peak complex functions,the following problems often occur： particles converge to the local optimal solution too fast, the late con verge slows down and search ability turns poor. In order to solve these problems, the definition of function complexity is proposed, and adaptive simulated annealing algorithm and tabu search algorithm are presented based on the function complexity. In these algo rithms, the step length control parameters are adaptively adjusted according to the function complexity;then rough solution of the function is obtained in terms of regulated step length;finally,and the original step length is used to acquire the global optimal solu tion. Experiments show that the proposed method cannot only transcend the limits of the local optimal solution, but also reduce item fion of the above algorithms,and efficiently improve local and global search ability.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2012年第6期1218-1222,共5页
Acta Electronica Sinica
基金
国家自然科学基金项目(No.61072109
No.61142011)
中央高校基本科研业务费专项资金(2012年度)
西安市科技局计划项目(No.CXY1133(1)
No.CXY1119(6))
关键词
函数复杂度
模拟退火算法
禁忌搜索算法
函数优化
function complexity
simulated annealing algorithm
tabu search algorithm
function optimization
