Using data-fuzzification technology in small data set learning to improve FMS scheduling accuracy

Production decisions in real dynamic flexible manufacturing systems (FMS), especially in the early stages are often made with limited information. Information is limited because scheduling knowledge is hard to establish in such an environment. Though the machine learning technique in the field of Artificial Intelligence is thus used for this task by many researchers, this research is aimed at increasing the accuracy of machine learning for FMS scheduling using small data sets. Approaches used include data-fuzzifying, domain range expansion, and the application of adaptive-network-based fuzzy inference systems (ANFIS). The results indicate that learning accuracy under this strategy is significantly better than that of a traditional crisp data neural networks.