Abstract: To improve the accuracy of demand forecasting,a multi-variable gray demand forecasting model by addressing customer perception factors was constructed.The modified gray relational analysis was adopted to analyze the importance of customer perception factors such as quality,performance and service.Meanwhile,the initial predicted value of each factor obtained by combining GM(1,1,x(0)) and weighted Markov model to be the input of GM (1,N,x(0)) which was employed to forecast the future demand trends.The example of demand forecasting for engine products demonstrated the applicability and validity of the proposed approach.

Key words: customer perception factors, modifiedgray relational analysis method, gray model, GM(1,N,x(0)), weighted Markov prediction model, demand forecasting

摘要： 随着顾客主导型市场的发展,顾客满意度对未来市场需求的影响日益显著,为提高需求预测的准确性,构建了考虑顾客感知要素的多变量灰色组合需求预测模型。引入改进的灰色关联分析方法对与需求变化有关的质量、性能、服务等顾客感知要素的重要度进行分析。对由单变量灰色模型的派生形式GM(1,1,x(0))得到的各自变量因子的初始预测值进行加权Markov模型修正,以确定多变量灰色模型的派生形式GM(1,N,x(0))的输入量,预测未来需求趋势。以某企业发动机产品的需求预测为例,证明了该方法的适用性和有效性。

关键词: 顾客感知要素, 改进的灰色关联分析, 单变量灰色模型, GM(1, N, x(0)), 加权Markov模型, 需求预测