Study on multi-task oriented services composition and optimisation with the ‘Multi-Composition for Each Task’ pattern in cloud manufacturing systems

Abstract

Recently, cloud manufacturing has been generating a great deal of interest among both practical users and researchers. Multi-task oriented manufacturing cloud services composition and optimisation (MTO-MCSCO) is critical to the optimal allocation of manufacturing resources and capabilities in cloud manufacturing systems. However, if users’ QoS requirements on multi-functionality manufacturing tasks (MFMTs) are high enough, no competent composite services can be identified based on the concepts of the single-task oriented services composition and optimisation (STO-SCO) technique, which was previously in use, and the currently existing ‘Each Composition for Each Task’ (ECET) pattern. To circumvent this, a ‘Multi-Composition for Each Task’ (MCET) pattern based global approach is proposed to combine the incompetent composite services into a whole to perform each MFMT collectively, in order to ensure the success rate of QoS requirement fulfilment and the overall QoS outcome. This new issue of MTO-MCSCO with the MCET pattern is a more general problem than are the previous STO-SCO and the current ECET pattern. To formulate the problem, exterior aggregation patterns and formulas are proposed. To tackle the problem, a hybrid-operator based matrix coded genetic algorithm (HO-MCGA) is implemented. The experimental results indicate that the proposed MCET pattern based global approach significantly outperforms the previous approaches, and the proposed HO-MCGA is sound performance-wise.

Keywords:

• cloud manufacturing
• service composition
• quality of service (QoS)
• multi-task
• genetic algorithm (GA)

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 61203135), the National Key Technology R&D Program of China (No. 2012BAH19F01), the Natural Science Foundation Project of CQ CSTC (No. CSCT2012JJA40020), and the Fundamental Research Funds for the Central Universities (No. CDJZR12180013).