一种灰色预测模型的新方法  

A New Method for Grey Forecasting Model

作　　者：谢开贵[1] 何斌[1] 谭界忠[1] 杨万年[1] 

机构地区：[1]重庆大学系统工程及应用数学系

出　　处：《系统工程理论与实践》1998年第7期69-75,共7页Systems Engineering-Theory & Practice

摘　　要：给出了一种灰色预测模型的新方法。该方法运用线性回归对数据进行分组，并对各组数据进行指数插值；运用二次规划进行权重选择；运用灰色灾变预测确定权重的时序集，从而求得预测值。这种方法提高了预测精度，并反映了事物波浪式前进发展的本质，拓广了灰色预测的范围。In this paper, a new method for grey forecasting model is proposed.The method applied linear regression to data classification, exponential interpolation to every group, quadratic programming to select weights and grey calamity forecasting to acquire the time series set of weights.It improved the forecasting precision and conformed to the development law of advancing in a series of wave so as to expand the application field of grey forecasting model.A practical example shows the effect of the method is remarkable.

关 键 词：灰色预测 线性回归 权重系数 灰色系统 

分 类 号：N94[自然科学总论—系统科学]