基于灰关联分析的时间序列改变点搜索研究——以华东地区农业受灾面积序列为例

Change-point Searching of Time Series based on Grey Relational Analysis ——Taking Sequences of Agricultural Disaster Area in East China as an Example

采用灰关联分析方法,构建了短时间序列中的改变点搜索算法,有效地解决了短时间序列改变点寻找的缺陷。首次将灰关联度算法应用到气象灾害时间序列改变点的搜索中,利用华东地区农业受灾面积时间序列实例来验证灰关联度算法的实用性和有效性。结果表明,用灰关联方法可以有效地获得改变点,并以此对时间序列的发展规律进行趋势与规律分析。

An algorithm is constructed with grey relational analysis to detect change-points in short time series, which effectively overcomes shortages of the existing statistical methods. The grey relational algorithm is firstly applied to the change-point searching of time series in meteorological disasters. Sequences of agricultural disaster area in East China are used to show validity and practicability of the algorithm of new grey relational method. The results indicate that the grey relational algorithm can effectively get the change-point. And in this case, the trends and laws of time series characters can be developed.