摘要
针对人工蜂群算法搜索效率低、易陷入局部最优和精度低等缺点,提出混合蜂群(hybrid bee colony,HBC)算法。将人工蜂群(artificial bee colony,ABC)算法局部收敛性与模拟退火(simulated annealing,SA)算法全局收敛性结合,为ABC算法提供了一种新机制。根据SA算法中Metropolis接受准则,通过调整温度依概率确定全局最优解的替代值,并利用全局最优解的替代值和个体极值来改进ABC算法的引领蜂搜索模式。其次,改进侦察蜂搜索方式,根据迭代次数非线性减小侦察蜂搜索范围和以一定概率反向搜索更新方式,能够有效地提高算法的全局搜索能力,并加快算法的后期收敛速度。通过对8个复杂函数仿真测试,结果表明,HBC算法在搜索性能和精度方面均有明显提高。
The hybrid bee colony (HBC) algorithm is proposed to overcome the problems that the artificial bee colony (ABC) algorithm is easier to get local optimal solution and has poor local search capability and low accuracy. The combination of ABC algorithm's local convergence and simulated annealing (SA) algorithm's global convergence provides a new mechanism for the ABC algorithm. The algorithm integrates the information of concept of the global best solution's the replacement value and individual optimum into the search equation of ABC algorithm. The selection method of substitution value is implemented through the Metropolis criterion of SA algorithm, namely, by regulating the temperature in probability to determine the replacement value of the global best solution. Then, according to the nonlinear to reduce the number of iterations to reduce the scouts search, so the search of the scout bees is also improved. This improvement can effectively improve the global search ability of the algorithm, and accelerate later convergence of the algorithm. The simulation results in eight complex numerical functions show that the proposed algorithm has better performances than traditional algo- rithms in search ability and accuracy.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2014年第10期2094-2101,共8页
Systems Engineering and Electronics
基金
中国国家自然科学基金委管理学部重点项目(71231001)
中央高校基本科研业务费专项资金(FRF-SD-12-020A)资助课题
关键词
人工蜂群算法
模拟退火算法
混合蜂群算法
函数优化
artificial bee colony (ABC) algorithm
simulated annealing (SA) algorithm
hybrid bee colony (HBC) algorithm
numerical function optimization