复杂动力巨系统中子系统行为间相关性研究的一种新方法

New Approach to the Correlation Measurement for Subsystems in a Complex Giant System

复杂开放巨系统内部关系错综复杂并具有动态特征,提出了非线性相互预测算法,分析一类复杂巨系统的内部多个子系统之间行为的相互关系及表征其非线性依赖关系的耦合强度.该方法在相空间重构的基础上,仅通过小数据和微信号可得到表征子系统间依赖关系的判别值,进而通过对判别值的分析得到了各子系统的行为特征及其相互作用关系.同时文中得到的子系统间的相互作用机制,可以为金融危机的理论分析提供非线性相互预测度量.

A nonlinear mutual prediction approach is presented to investigate the correlations and coupling strengths of nonlinear dependence among the subsystems in an open complex gi- ant system, which behaved with complicated nonlinear dynamical characteristics. The time-var- ying discriminant values of mutual dependence obtained with the proposed method could be used to predict the correlations of the subsystems in a giant system based upon phase space re~ construction by using the observed small data and micro signal. Moreover, the obtained mechao nlsm of interaction between the subsystems provides a nonlinear mutual prediction measure- ment, which is suitable for the analysis of financial crisis analytically.