A Genetic Algorithm-Based Model for Inventory Control in Intermittent Demands

Abstract

Demand forecasting is a difficult field of study for intermittent demands. Spare parts demand structures also have an intermittent demand structure. Therefore, for companies operating in this field, this situation causes various problems (holding cost or cost of lost sale). Intermittent demands are inherently difficult to predict. Demands with a smooth structure provide a more suitable working environment for businesses. Because the more accurately the relevant demand is forecasted, the more smoothly the works that depend on demand forecasting are carried on. In this study, a randomly generated demand series with intermittent demand structure is examined. The estimation difficulty of intermittent demand is illustrated by an estimation made in Matlab. In order to avoid this difficulty, the costs were tried to be minimized by determining the inventory levels. An inventory model is proposed that determines stock levels using intermittent demands and calculates average profit by calculating costs. The related model was solved with Genetic Algorithm in Matlab and the results were recorded.