弱化缓冲算子对GM(1,1)模型的预测效应及适用性

在GM（1，1）模型预测中，缓冲算子能有效处理含有冲击扰动因素的原始数据，改善模型的预测效果. 本文在系统分析缓冲算子对GM（1，1）预测作用过程的基础上，提出了GM（1，1）模型的预测效应以及缓冲算子适用性的评价准则. 选用典型的6种弱化缓冲算子对河南粮食产量数据分长序列、宽间距序列和短序列三种情况进行了模拟计算分析. 确立了不同算子对GM（1，1）模型预测产生的效应及其适用范围，并选用具有类似特征的其他数据序列验证了研究结果的有效性.

Forecasting effect and applicability of weakening buffer operators on GM(1, 1)

In GM(1, 1) prediction, buffer operators can effectively process raw data with disturbance factors and improve the effect of GM(1, 1) prediction. In this paper, the forecasting effect of GM(1, 1) and applicability evaluation criteria of weakening buffer operators were proposed based on systemic analysis of buffer operators working process to GM(1, 1) prediction. It conducted simulation analysis by choosing six kinds of typical weakening buffer operators working on grain production data of Henan province which were divided into three sequences (the long sequence, the wide spacing sequence and the short sequence). Finally, this paper establishes the effect and its domain of application of each buffer operator on GM(1, 1) prediction. Meanwhile, the validity of the conclusions is verified by selecting other data sequences with similar properties.