Prediction of Lateral Deformation of Diaphragm Wall Support Structures in the Deep Foundation Pit by Using Improved Exponential Smoothing Method

Due to a large number of uncertainties in the excavation process of the foundation pit,the real-time deformation reliability analysis and prediction of its supporting structure can not be carried out according to the changes of the site environment in the design stage.The distribution range of enclosure structures is large,and the failure location is unpredictable.Due to the limitation of the number of monitoring points,the engineering monitoring data cannot directly reflect the continuous change of the retaining structures.Based on the method of up-crossing rate,was proposed for deep foundation pit engineering with an underground diaphragm wall.An adaptive exponential smoothing model is established by using MATLAB to automatically calculate and predict the lateral deformation of the deep foundation pit retaining structure at different depths.By comparing the measured and predicted lateral deformation data,the results show that the self-adaptive exponential smoothing model has the characteristics of higher accuracy and smaller error fluctuation in short-term field prediction,the transcendental probability analysis based on the prediction data shows that the analysis model based on the lateral deformation of deep foundation pit is also applicable to the measured data obtained by the self-adaptive exponential smoothing method,and the analysis accuracy will be improved as the displacement limit is far away from the allowed displacement.The inverse distance weighted interpolation method based on ellipse search mode solves the defect that the sample points are distributed in the depth direction,which makes the interpolation result smoother.Combined with the analysis model based on the lateral deformation of the deep foundation pit,the deformation prediction of the spatial surface of the retaining structure is performed,which makes the evaluation method more intuitive and concise.