Parameter estimation and bad data identification of grey systems

Chongqing Kang,Qing Xia,Nian de Xiang
1997-01-01
Abstract:The methods of parameter estimation and bad data detection and identification are applied in grey system modeling to improve its accuracy. The proposed methods can be used as data preparation techniques to detect bad data and fill the blank with a more accurate value in an initial series. The modeling accuracy of the grey system is greatly improved by such methods. As an application, power system load forecasting using grey models is studied and some more reasonable forecasting results are got comparing to conventional grey system methods.
What problem does this paper attempt to address?