灰色系统理论中二维数据序列均值生成的讨论

灰色系统理论中非等时序序列或含有空穴的序列,通常可以用均值生成方法来填补,生成等时序序列。本文研究发现,用非紧邻均值生成方法生成的内点与实际存在有明显的误差,为此讨论了均值生成和有偏生成算法。利用数据序列点的局部凹凸特性来表示数据序列点的特征,提出了采用数据序列点连线的状态选取生成系数的方法。通过对一维数据序列传统均值生成方法的比较,给出了二维数据序列的定义,讨论了在二维数据序列下的均值生成方法,提出了有偏生成算法中生成系数分解的选取方法。

DISCUSSION ON AVERAGE GENERATION OF TWO-DIMENSIONAL DATA SEQUENCE IN GREY SYSTEM THEORY

An unequal time interval sequence or a sequence with blanks is usually completed with average ge-neration in grey system theory. This paper discovers that there exists obvious errors when using average generation to generate internal points of non-consecutive neighbours. The average generation and the preference generation of the sequence are discussed, the concave and convex properties show the status of local sequence and propose a new idea for using the status to build up the criteria of choosing generation coefficient. Compared with the general average method of the one-dimensional data sequence, the two-dimensional data sequence is defined and its average generation is discussed, and the coefficient decision method for the preference generation is presented.