An optimal combination prediction method of turnover spare parts consumption based on certain weight

Abstract

Abstract Aimed at the situation that the first, second and third exponential smoothing method and the three and four term moving average method are used to forecast the consumption of spare parts separately, the combined prediction model is established by optimizing the combination of these single prediction methods, and the result with the minimum prediction error is taken as the predicted value of consumption. A concrete example shows that the combined prediction model is more accurate and has higher practicability and generalization.