不同趋势对滑动移除近似熵的影响  

Effects of different trends on moving cut data-approximate entropy

作　　者：金红梅[1] 何文平[2] 侯威[2] 章大全[2] 

机构地区：[1]兰州大学大气科学学院,兰州730000 [2]国家气候中心,北京100081

出　　处：《物理学报》2012年第6期501-509,共9页Acta Physica Sinica

基　　金：全球变化重大研究计划(批准号:2012CB9552037);国家自然科学基金(批准号:40930952;41175067和40905034);国家科技支撑计划(批准号:2007BAC29B01;2009BAC51B04)资助的课题~~

摘　　要：许多观测资料中存在着各种各样的趋势,如季节变化引起的周期性趋势、全球变暖所造成的线性趋势、多项式趋势等.鉴于此,本文研究了各种趋势对滑动移除近似熵的影响.数值试验结果表明,周期性趋势、线性趋势及非线性趋势均对滑动移除近似熵的突变检测结果影响较小,论证了滑动移除近似熵方法检测突变的可靠性,为该方法在实际观测资料中的广泛应用提供了实验基础.Various trends exist in many observation data,such as periodical trend caused by seasonal variation,linear trend and polynomial trend brought about by global warming.In the present paper,the effects of different trends on moving cut data-approximate entropy are investigated.The numerical tests on model time series indicate that the detection results of moving cut data-approximate entropy are little affected by periodical trend,linear trend and nonlinear trend.The reliability of abrupt change detection of moving cut dataapproximate entropy is demonstrated,which provides an experimental basis for the wide applications of the present method in real observation data.

关 键 词：滑动移除近似熵 趋势 突变检测 

分 类 号：P416[天文地球—大气科学及气象学]