Fuzzy Learning System for Uncertain Remanufacturing Process Time of Used Components

In view of the uncertainties inherent in remanufacturing process, a fuzzy petri net based remanufacturing process model is proposed to explicitly represent the variations in remanufacturing process routings and process time with respect to product conditions. A fuzzy learning system with an adaptive learning mechanism is designed to estimate the remanufacturing process time of used component under specific conditions, and to dynamically reshape the distributions of actual process time to improve future predictions. The proposed method is illustrated through the remanufacturing of a batch of used lathe spindles with a Matlab-based simulation system.