大坝观测数据序列中的混吨现象

本文探讨大坝观测数据中的混沌现象。文中对坝体径向位移的实测值与残差的数据序列，分别进行相空间重构，将若干固定时间延迟点上的测量作为新维处理，形成相点，按照关联维数方法求算吸引子维数。计算中对数据进行规格化处理，所得吸引子维数为3.1～3.5.不同测点上的测值与不同起始时刻的计算结果比较接近与稳定。文中还绘制了实测值与残差序列的相轨线图，它们均与白噪声序列的轨线形状有明显不同。初步表明观测数据中存在有奇异吸引子。

Chaotic Phenomenon in Observation Data of Dam Monitoring

The chaotic phenomenon in observation data of dam monitoring is discussed.The data series of observed vaules and their residual errors are reconfigured in phase space respectively and the observed values at the points of certain time delay are handled as new dimensions.Thereby phase points are formed.The dimension amount of the attractor can be calculated according to the correlation dimension method.The calculated results shows that the dimension of the attractor obtained is in the range from 3.1 to 3.5.It seem that the phase traces in the plots of observed values and residual error series are different from white noise trace.These indicate that the strange attractor may exist in the observation data series.