摘要
设计一种新的混合蚁群算法,该算法以一种新的加权二进制蚁群算法为基础,将分布估计算法PB IL的概率分布模型用来指导蚂蚁路径的选择,同时对不同位置的蚂蚁采用加权系数来控制信息素散发量,根据信息素得到的转移概率、PB IL的模型概率及二者融合的概率来产生新的个体,保证了个体的多样性,从而提高了算法的快速性和全局最优解的搜索能力.通过测试函数优化表明该算法具有良好的收敛速度和稳定性,改善了蚁群算法容易陷入局部最优而早熟的缺陷.
A kind of new hybrid ant colony algorithm is designed. It uses a new binary ant colony algorithm with weight factor as foundation. Ants choose the route with the guidance of model probability of Population based incremental learning （PBIL） and control measure of pheromone emanating by weight factor. The new population are produced by probability model of PBIL, transfer probability of ants pheromone and theirs associative probability so that population polymorphism is ensured and the optimal convergence rate and the ability of breaking away from the local minima are improved. Optimization simulation resutts based on typical. functions show that the hybrid algorithm has the speedy convergence rate and stability.
出处
《数学的实践与认识》
CSCD
北大核心
2009年第6期154-161,共8页
Mathematics in Practice and Theory
基金
河北省自然科学基金(F2008001166)
关键词
蚁群算法
分布估计算法
PBIL
概率模型
Ant colony algorithm （ACA）
Estimation of distribution algorithm （EDA）
Population based incremental learning （PBIL）
Model probability
