多粒度制造服务建模与调度方法研究

Aiming at the characteristics of multi-level, multi-type, heterogeneity, autonomy and geographical distribution of manufacturing services, a unified description model of manufacturing resource was proposed including basic information, collaborative relationship, manufacturing capability, service quality and knowledge, and a multi-granularity manufacturing resource modeling method was developed including homogeneous and heterogeneous aggregation. Aiming at the share and efficient utilization of multi-level heterogeneous manufacturing resources, an improved K-means clustering-based multi-objective genetic algorithm (KGA) considering the requirements of time, cost, reliability and sustainability in manufacturing services aggregation was designed by adding elite sets and K-means clustering operator. The performance of multi-objective particle swarm optimization (MOPSO) and non-dominated sorting genetic algorithm Ⅱ (NSGA-Ⅱ) were compared with KGA on a complex product assembly cloud simulation platform. Experiment results show that through multi-granularity aggregation, manufacturing tasks and service resources can be matched and invoked more quickly, and the results are better on the service reliability and sustainability indicators.