基于函数cos（x^α）变换的灰色模型及其在变压器油中气体浓度预测的应用

Gray Model Based on cos（x^α） Transformation and Its Application in the Prediction of Gas Concentration in Transformer Oil

基于函数变换理论,采用函数cos(xα)(α≥1)变换建立了灰色预测模型,并从理论上证明该变换可以有效地提高离散数据序列的光滑度,且其光滑度比单纯的对数函数、三角函数、幂函数变换法更优。将建立的灰色预测模型用于变压器油中溶解气体浓度预测,其实例结果分析表明,该方法有效可行且预测精度较高。

The method of cos （x^α） （ α≥1 ） transformation is used to establish the grey forecasting model based on the transform theory of function. It is proved that the transformation can enhance the smoothness of the data effectively in theory, and more effective than the simple logarithmic, trigonometric or the power functions. The established gray model is used to predict the gas concentration in transformer oil. Example verification shows that the proposed approach can further enhance the precision of prediction effectively.