A fuzzy model for load-oriented manufacturing control

This paper presents an original approach to load-oriented manufacturing control for job-shop scheduling, based on fuzzy theory. The model allows to cope with the pitfalls encountered by traditional approaches to job-shop scheduling in the definition of system parameters. In fact, traditional approaches to job-shop scheduling assume that system parameters are deterministically known ex-ante; on the contrary, the parameters values actually observed in the job-shop are often different due to the impact of unforeseen dynamics. As a consequence, the effectiveness of traditional approaches is undermined. In this paper the authors focus on the ‘machine output in the planning horizon’ parameter and present a model allowing to represent that parameter as a neuro-fuzzy variable, whereas traditional approach represents it as a deterministic value. The case study carried out in a real manufacturing system and reported at the end of the paper shows the effectiveness of the proposed approach.