An intelligent algorithm for optimum forecasting of manufacturing lead times in fuzzy and crisp environments

Abstract

The manufacturing lead time (MLT) prediction is a vital activity in any manufacturing organisation. This study presents a comprehensive procedure for comparing fuzzy regressions (FR) and conventional regressions (CR), artificial neural network (ANN), adaptive network fuzzy inference system (ANFIS) and genetic algorithm (GA) for manufacturing lead time estimation in both crisp and ambiguous environments. According to a proper sensitivity analysis, the best model is selected based on the lowest mean absolute percentage error (MAPE). Weekly lead times for a large complex electric-motor assembly line are chosen as the actual case of this study. The results of the models implemented on the data for 70 weeks in the assembly line illustrate the applicability and superiority of the proposed algorithm.