周年极移序列的Markov链概型  

A MARKOV CHAIN PROBABLITY MODEL OF ANNUAL POLAR MOTION SERIES

作　　者：杨本有 孙长安 陶桂琴[1] 

机构地区：[1]中国科学院紫金山天文台,东南大学物理系

出　　处：《天文学报》1996年第4期368-376,共9页Acta Astronomica Sinica

摘　　要：本文根据Ｍａｒｋｏｖ随机过程理论，给出了周年极移振幅ｒａ序列的一阶Ｍａｒｋｏｖ链概型，并建立了ｒａ序列增加年Ｒ与减小年Ｄ游程以及（Ｒ十Ｄ）游程循环（周年极移循环）的概率分布模式，经统计假设检验，一阶Ｍａｒｋｏｖ链可以很好拟合。序列的变化过程．得到了ｒａ不同状态的平均首次到达时间矩阵，其值大多与实测符合．ｒａ的基本周期为２年、６．０７年、６．５４年、７．０８－８．５０年、２１．２５－２８．３３年和８５年，而６．０７年是它的一个主要周期．由Ｍａｒｋｏｖ链的遍历性定理得到，预报ｒａ时取用在此之前的第６—１３个以内的样本为宜．Based upon Markov's random process theory, a Markov chain probablity model of one order corresponding to the series of amplitute (ra) of annual polar motion is given. The probablity models of runs of R (increasing year) and D (decreasing year),and of their run cycle of (R+D), are also established. By means of the statistical hypothesis test, it seems that the Markov chain of one order fits to the change process of ra series. The obtained matrices of average arrival times of different r'a s show that its basic peroids are 2, 6.07, 6.54, 7.08-8.50, 21.25-28.33 and 85 years,and among them 6.07 years is the main one. From the ergodic theorem of Markov chain, the suitable samples are taken from the 6th to 13th points before the predicted point.

关 键 词：周年极移 马氏链概型 游程 极移 地球自转 

分 类 号：P183.31[天文地球—天文学]