A New Optimal Demand Forecast Model

This paper presents a model for determining an optimum demand plan which has been developed whereby both demand figures from the past as well as the results of market research concerning price elasticity are known. The sales demand demonstrates a high seasonal fluctuation as well as a shortterm fluctuation, which is also relevant. The production capacity is restricted . The main idea of the model is first of all to approximate the past demand figures in order to predict future demand figures through extrapolation. In the second step, a non-linear production optimisation model is applied for calculation of the cumulative demand plan taking into account both the price elasticity and resource restrictions. The fmal predicted demand data is shown by an extrapolated curve corrected by a constant. The latter is determined by the fact that the integral of the demand curve has to be the cumulative demand plan. In addition to this result, a beta level of service and consequently the level of the safety stock can be calculated by means of a curvilinear regression analysis .