需求预测和供应链协调对长鞭效应的影响分析

运用自适应过滤、滑动平均和指数平滑预测方法分析多阶段供应链中长鞭效应．分析结果表明，在运用自适应过滤预测方法时，供应链不一定存在长鞭效应，文章进一步提出该预测方法前提下长鞭效应存在的条件；在运用滑动平均和指数平滑预测方法时，长鞭效应总是存在．根据三种预测方法的分析结果得出，适当的预测参数和供应链协调可以减小长鞭效应．并在此基础上，比较分析三种预测方法条件下供应链协调时需求参数对长鞭效应的影响规律，结果表明，预测方法不同，需求参数对长鞭效应的影响规律也有所不同．

Impacts of demand forecasts and supply chain coordination on bullwhip effect

In this paper, the bullwhip effect in a multistage supply chain is analyzed using self-adaptive filtering, moving average and exponential smoothing demand forecast technniques. We find out that the bullwhip effect does not always exist when the self-adaptive filtering forecasting is used, and present the existence condition. However, the bullwhip effect always exists when the moving average and the exponential smoothing forecasting are used. Based on these three forecasting tools, it is concluded that suitable forecast parameters and supply chain coordination will alleviate the bullwhip effect. In addition, we compare the change behaviors of the bullwhip effect with respect to demand parameter under three different forecasting scenarios when a supply chain is coordinated. Our analysis shows that these change patterns of the bullwhip effect are different due to different forecasting tools.