Application of Empirical Mode Decomposition (EMD) in Processing Data of the Slope Deformation

摘要：

The slope is a complex system whose deformations show complex nonlinear behavior. The time series from its deformation shows obvious nonlinear and nonstationary characteristics. The traditional time series analysis method is based on the stationary series so it is difficult to improve the precision of forecasting future deformation. In this paper the method of empirical mode decomposition (EDM)combining the radial basis function (RBF) neural network is discussed. In this method, firstly, the method of empirical mode decomposition is used to decompose original series to get the intrinsic mode functions (IMF) and the trend part, which are the stationary series; secondly, the RBF neural network is used to fit each IMF for forecasting future data; finally, again the RBF neural network is used to fit together each result forecasted and the trend part to get the final result The experiment results indicate that this approach is effective to nonlinear and nonstationary time series.

关键词： nonlinear time series empirical mode decomposition (EMD) intrinsic mode functions (IMF) neural network

作者: Xu Jia Ma Fenghai Huang Shengxiang Yang Fan

作者单位: School of Geomatics, Liaoning Technical University, Fuxin 123000 China Dalian University, Dalian 11662, China School of Geodesy and Geomatics Wuhan University, Wuhan 430079 China

会议类型: 国际会议

会议名称: The 3rd International Symposium on Modern Mining & Safety Technology Proceedings(第三届现代采矿与安全技术国际学术会议)

会议地点: 辽宁阜新

会议语种:英文

页码: 817-820

在线出版日期: 2008-08-04（万方平台首次上网日期，不代表论文的发表时间）