基于滑动平均的灰色GM(1,1)模型在水泥产量预测中的应用

Moving Average Based Grey Model GM(1, 1)for Cement Output Forecast

本文介绍了基于滑动平均的GM（1，1）模型的建模过程，并分别采用传统灰色系统理论的GM（1，1）模型和基于滑动平均的GM（1，1）模型，用MATLAB语言编程，对我国1991～2007年的水泥产量进行拟合预测，并对预测结果进行了残差检验、关联度检验和后验差检验。通过比较分析，结果表明，基于滑动平均的GM（1，1）模型能够减少数值的过度波动，预测精度较高，具有更高的实用性，可为政府有关部门和相关企业做出正确决策提供科学依据。

The modeling procedure of GM（1,1） model based on slip average is introduced. According to the cement output of China from 1991 to 2007, using traditional GM （1,1） of Grey System theory and GM（1,1） model based on slip average distinguishingly and programming by MATLAB, the cement output is fore- casted. With the forecast results of these models, the residual error, relevancy degree and error inspection of post-sample are also tested. By Comparison and Analysis, the result shows that, the GM （1,1） model based on slip average can decrease the extravagant fluctuation of the input values and presents higher precision and better practicability, so the future output forecasted by this model can help correlative departments and enterprises with proper decision-making.