Application of an adaptive neuro-fuzzy inference system in inventory control

A model based on adaptive neuro-fuzzy inference systems is proposed for inventory control. The periodic review model of inventory control with variable order quantity is considered. The model takes into account the dynamics of production-inventory system in a control theoretic approach. The control module combines fuzzy logic and the conventional proportional-integral (PI) algorithm. The fuzzy logic membership functions are adapted to fit the training data set of production-inventory. It simulates the decision support system to maintain the inventory, of the finished product, at the desired level, in spite of variations in demand. The effectiveness of the proposed control model is illustrated using the real data of a typical packaging organization operating in the Sultanate of Oman.