基于离散人工群算法的云制造服务组合

摘要

With the rapid development of network technology such as Internet,cloud computing and so on,single manufacturing service has already not satisfied the increasingly complex tasks for users. So,cloud manufacturing service composition,as a NP hard problem,has been the applied and research hotspot in recent years. As to service composition optimal selection,a Location Search Discrete Artificial Bee Colony ( LSDABC) is proposed in this paper based on improvement of original ABC to provide the service composition execution path with optimal QoS for users. This algorithm introduces selection probability based on population and local search strategy to improve the exploitation ability and convergence speed and to avoid falling into local optimum. Finally,LSDABC is applied to the cloud manufac-turing service composition. The experiment shows that the LSDABC has better quality and robustness compared with the original ABC, DE and PSO.%随着互联网、云计算等网络技术的快速发展,单一制造服务已无法满足用户日益复杂的制造任务,所以云制造服务组合问题一直是近年来应用和研究的热点,为典型NP难题。文中针对云制造服务组合优选问题,改进原始人工蜂群算法( Artificial Bee Colony,ABC),提出了一种基于局部搜索离散蜂群算法( Location Search Discrete Artificial Bee Colony,LSD-ABC),从而为用户选择服务质量( Quality of Service,QoS)最优的服务组合执行路径。该算法引入种群的选择概率和对最优解的局部搜索策略,提升算法的开采能力、收敛速度,同时避免出现搜索停滞陷入局部最优。最后将LSDABC应用于云制造服务组合优选中进行仿真实验,并将结果与原始ABC、DE、PSO算法进行对比。实验结果表明,LSDABC具有较好的求解质量和鲁棒性。