EXPLORATION OF THE BULLWHIP EFFECT BASED ON THE EVOLUTIONARY LEAST-MEAN-SQUARE ALGORITHM

The bullwhip effect, a well known phenomenon occurring in business activity where the demand information is not fully shared among the members of supply chain, conducts the upstream manufacturer to excessively anticipate the demand capacity of the downstream retailer. The manufacturer improperly decides the amount of the products not only to raise the inventory cost on the way of poorly handling the actually downstream demand, but also to lose the chance of business deals due to its backordering. To cope with the bullwhip effect by taking into account the holding and backorder costs, this paper proposed a prediction system based on an evolutionary least-mean-square algorithm to estimate the downstream demand, which consequently enables the batch ordering of manufacturer to close the estimated inventory level.