摘要
本文分两个部分分别探讨了单一和两维时间序列谱分析在研究作物量变过程规律中的意义、应用的原理和方法,并通过实例结果的分折表明这些方法的可行性,所揭示的生物学意义和这些方法的应用在研究作物量变过程中将可能具有的广泛的研究前景。
In this paper spectrum analysis of time series has been applied in the study of the dynamic processes of upland rice variety 'IR36' of leaf, root and panicle weight, By means of calculating following functions the three dynamic Processes has been studied. 1. Autocorrelation function: R(t)=1/N-t sum from i=1 to N-t ·,=xi-x/δ 2. Power spectrum: (K)=1/M[R(0)+2 sum from t=1 to M-1 R(t)·cos kπt/M+R(M)·cos kπ] 3.Cross-correlation functions: Rab(t)=1/N-t sum from i=1 to N-t , Rba(t)=1/N-t sum from i=1 to N-t . 4.Cross-spectrum (coherncy spectrum): vab=C_(ab)~2(K)+Q_(ab)~2(K)/S_a(K)·S_b(K) etc. The conclusions are that although there are similar weighted fashions between leaf and root, there are different relative reactions between leaf to panicle and root to Panicle, and that there is another relative reaction between root to leaf. So the accumulation of the panicle weight is a complex relative reaction process. In order to get higher panicle yield, it may be necessary to get oPtimie dynamic accumulative Processes of leaf, root, stems and panicle weight.
出处
《生物数学学报》
CSCD
北大核心
1990年第2期188-192,共5页
Journal of Biomathematics
关键词
时间序列分析
作物量变规律
谱分析
time series analysis, spectra analysis, random process, upland rice, quantitative changing rules of the crop.
