一种多层次制造服务建模和组合优选方法

为提高云制造环境下服务建模和组合优选的准确性，首先将制造服务分多个层次进行描述，从资源服务、功能服务和流程服务3个层次进行建模。然后针对多层次服务模型，采用服务执行时间、服务花费成本和服务用户评价等因素构建服务组合优选的质量评估函数。为解决多层次服务的组合优选问题，提出一种改进引力搜索算法（NGSA），将小生境中的拥挤度因子和适应值共享技术引入传统引力搜索算法（GSA）以提高收敛速度和准确性。算例验证表明，相比传统的遗传算法（GA）和粒子群优化（PSO）算法，NGSA能在较短的时间内收敛，且最优解的匹配准确度更高。 

A method of multi-level manufacturing service modeling and combinatorial optimal-selection

In order to improve the accuracy of service modeling and combinatorial optimal-selection in cloud manufacturing, a multi-level modeling methodology is proposed to describe manufacturing services, which subdivided the service into three fine-grained levels:resource service, function service and process service. From the perspective of QoS indexes, the relationship among execution, time service cost and user evaluation for different service levels are analyzed and elaborated, and the corresponding evaluation objective functions of services composition are established. A niching behavior based gravitational search algorithm (NGSA) is designed to address manufacturing services composition problem, in which the niche crowding factor and fitness sharing technology are applied to gravitational search algorithm (GSA) to improve its convergence speed and accuracy. Finally, the simulation research results demonstrate that the NGSA algorithm can search better solution with less time-consumption than the traditional algorithms such as genetic algorithm (GA) and particle swarm optimization (PSO) algorithm.