Modeling of service correlation for service composition in cloud manufacturing

Abstract

Cloud manufacturing (CMfg), integrating distributed manufacturing resources as services to cloud center, aims at intelligent, green, and economic customized manufacturing. The optimal composition of services to fulfill particular manufacturing requirement is a core issue to realize efficient cloud manufacturing. Many researchers have studied the problem considering the Quality-of-Service (QoS) of independent services. However, the correlation between services is rarely considered. In this paper, the importance of service correlation is emphasized. Two kinds of service correlation, service exclusion and service collaboration, are modeled for service composition. An improved algorithm DET, which combines Differential Evolution Algorithm (DE) with a Tabu table based on service exclusive and collaborative relationships, is designed to filter composable services and find better solutions for complex tasks. Experiments have shown the effects of service correlation on the quality of composed services and demonstrated the effectiveness of the proposed method DET compared with traditional DE.