桥梁变形数据小波去噪与混沌预测  

WAVELET DENOISING AND CHAOS PREDICTION OF BRIDGE DEFORMATION DATA

作　　者：栾元重[1] 栾亨宣[2] 马德鹏[3] 刘娜 周岩[1] 

机构地区：[1]山东科技大学测绘科学与工程学院,青岛266590 [2]山东科技大学机械电子工程学院,青岛266590 [3]山东科技大学资源与环境工程学院,青岛266590 [4]重庆地矿测绘院,重庆400042

出　　处：《大地测量与地球动力学》2013年第5期133-135,139,共4页Journal of Geodesy and Geodynamics

摘　　要：跨海大桥因海浪、飓风、船体撞击等外界因素作用,桥梁变形表现为动态非线性变形状态,也存在混沌现象。为此,选择db10小波和软阈值消噪法分解与消除变形数据中的粗差与噪声,并对小波预处理后的平滑变形时间序列通过相空间重构,运用Lyapunov指数进行桥梁混沌状态判定;再将混沌时间序列预测结果与指数平滑法预测值进行分析对比,证明混沌预测可靠且精度高。Due to external factors such as waves, hurricanes and ship collision, crosssea bridge deformation shows the dynamic nonlinear state, there is also a chaotic phenomenon. Therefore, db10 wavelet threshold denoising and soft law are used to break down and eliminate the gross errors in the data distortion and noise. Phase space is reconstructed by the smooth deformation of the timeseries after wavelet pretreatment, the bridge chaos state is determined by Lyapunov indices, and chaotic time series prediction and exponential smoothing forecast are analyzed and compared, it is proved that chaos prediction has reliable and high precision.

关 键 词：软阈值消噪法 相空间重构 LYAPUNOV指数 混沌时间序列 指数平滑法 

分 类 号：P207[天文地球—测绘科学与技术]