Slope deformation prediction method considering uncertain influence of different elements

• Main Authors: XIE XIUDONG, FU JIAYUAN, LIN CHUAN, ZHENG ZHIMING, HUANG BIN, SU YAN, LAI XIAOHE, CHEN YAOXIN
• Format: Patent
• Language: Chinese; English
• Published: 25.04.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention relates to a slope deformation prediction method considering the influence of different-element uncertainty. The method comprises the following steps: processing related historical monitoring data; dividing the processed data into a learning sample set, a calibration sample set and a prediction set; constructing a plurality of interval prediction models by taking the learning sample set as the input of the anisotropic elasticity-based long-short-term memory neural network; combining the obtained interval prediction model with a corresponding calibration sample to perform preliminary perception, obtaining a corresponding shape-preserving score group according to a relationship between an actual observation value and an output prediction interval, and calculating empirical probability quantiles respectively; and performing adaptive optimization on the prediction interval generated by each model by using an empirical probability quantile, so that each constructed interval prediction model obtains o