以混沌理论为基础,提出了河流混沌特性分析方法。选择对河流演变有重要影响的宽深比时间序列和水沙时间序列,首先对这些时间序列进行相空间重构,计算不同河型的宽深比、径流量和含沙量时间序列的饱和关联维数和最大Lyapunov指数,然后通过求这些时间序列的饱和关联维数的加权平均值和最大Lyapunov指数的加权平均值,得出不同河型的混沌特性。以黄河下游的6个河段3种河型为例,对宽深比、径流量和含沙量时间序列,分别进行混沌特性分析。研究结果表明,河流演变具有明显的混沌特性,但不同河型表现出的混沌特性不同,游荡河型混沌特性较强,弯曲河型混沌特性较弱。通过对河流混沌特性分析,有助于加深对河流演变预测的进一步认识。根据混沌理论,混沌系统短期行为可以预测,而长期不能预测。所以,河流演变预测是短期可行,长期很难预测、甚至是不可预测的。
Based on chaos theory, the method of identifying the chaotic characteristics of channel pattern is proposed. An important effect is studied on the river evolution by time series between width-depth ratio and flow-sediment. Firstly the phases space reconfiguration are completed about these time series. Then the satu- rated relation dimension and the maximum Lyapunov exponent of these time series are obtained. The weighted average of saturation correlation dimension and the weighted average maximum Lyapunov exponent of these time series indicate chaotic characteristics of channel pattern. Chaos characters of width to depth ratio time se- ries, runoff time series and sediment concentration time series for 6 reaches ( including 3 channel patterns) of the lower Yellow River are analyzed. Results show that, there are obvious chaotic characteristics in river evolu- tion, but different channel pattern has different chaotic characteristics, chaos characteristics of wandering river is more stronger than meandering river. According to chaos theory, chaotic systems can be predicted in the short-term, not predictable in the long-term, therefore river evolution only can be predicted in the short-term.