Parameter Identification of Non-equigap Grey Prediction Model under Multiple Transformation
The data prediction and control problems of large quantity of non-equigap series exist in complex systems, the non-equigap grey prediction control model is a kind of dynamic model described by a differential equation, which is applicable to original series with unequal sampling intervals and sharp variation. Based on the mathematical modeling theory of poor and uncertain information system, in this paper we firstly research the quantitative relation between the multiple transformation of the original sequence and the parameters of the non-equigap GM(1,1) model, and obtain its parameter identification and forecasting errors under multiple transformation. Then we propose the optimization calculation and extending applications by choosing appropriate multiple transformation to the original sequence according to the actual figures. At last an empirical example indicates that the method is not only simple and effective, but can also resolve some ill-conditioned problems of the non-equigap GM(1,1) model.
Xinping Xiao Fuqin Li
School of Science Wuhan University of Technology Wuhan, 430063, Hubei, China
国际会议
南宁
英文
2007-07-20（万方平台首次上网日期，不代表论文的发表时间）