并行二进制蚁群算法的多峰函数优化  

Parallel Binary Ant Colony Algorithm for Multi-modal Function Optimization Problems

作　　者：王柳毅[1] 熊伟清[1] 

机构地区：[1]宁波大学计算机科学与技术研究所,宁波315211

出　　处：《计算机工程与应用》2006年第22期42-45,共4页Computer Engineering and Applications

基　　金：国家自然科学基金资助项目(编号:60272099)

摘　　要：针对已有蚁群算法在函数优化问题上存在的几个不足:如算法实现较难,占用过多的存储空间,需要记忆功能,不容易与其他算法结合等等,提出了二进制蚁群算法。实验证明该算法在处理单极值问题时有较好的表现,但是在处理多峰函数时存在着一定的缺陷,对此,论文对该算法进行了改进,将并行化引入算法。通过对几个函数的测试(包括多峰和单峰),结果表明该改进算法具有较好的稳定性和收敛速度,算法性能良好。Aiming at solving the drawbacks of the original Ant Colony Algorithm on function optimization problems： implementation of the algorithm is difficult,too much storage room is needed,memory function is involved,and it is not convenient to combine the algorithm with others etc,we propose Binary Ant Colony Algorithm（BACA）.Experiment result shows a good performance when the algorithm is applied in single-modal function optimization problems,however,when coming across the multi-modal function,the algorithm still has several shortages,therefore,this paper modifies the algorithm,and introduces parallelization to it.After the algorithm is applied to several testing functions,the result shows a good convergence speed and stability,and the performance of the algorithm is fine.

关 键 词：蚁群算法 二进制 函数优化 并行化 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]