关键词: 滑坡变形, ELM模型, 去噪分析, 误差补偿, 丹巴滑坡,

Abstract: In order to achieve high-precision prediction of landslide deformation,taking Danba Landslide as an example,using the new intelligent ELM model as the theoretical basis,an error compensation intelligent prediction model for landslide deformation prediction was constructed by de-noising analysis,parameter optimization and error compensation prediction.The results showed that both wavelet de-noising and Kalman filtering can remove the error information from the deformed data very well,and the wavelet de-noising effect was relatively better.At the same time,on the basis of de-noising data,the prediction accuracy can be effectively improved by optimizing the parameters of the prediction model and weakening the error,the average relative error was less than 2%,which verified the validity of the prediction model.Moreover,the model had high prediction accuracy in mid-term and late-stage deformation prediction of landslide.It further showed that the model not only had good prediction accuracy,but also had high reliability,and was suitable for landslide deformation prediction research.In this paper,a new way was provided for the analysis of landslide deformation trend and law,which can guide the prevention and control of landslide in the later stage.

Key words: landslide deformation, ELM model, de-noising analysis, error compensation, Danba Landslide,