A landslide displacement prediction method based on a hybrid machine learning model

Abstract

The invention discloses a landslide displacement prediction method of a hybrid machine learning model. The landslide influence factor mining and landslide displacement data acquisition and preprocessing are performed on the landslide monitoring points, and the data is used as the training data of the model; the obtained landslide influence factors and landslide displacement data are obtained. It is regarded as a time series for wavelet de-drying processing; based on the principle of time series, Hodrick-Prescott filter (HP) is used to decompose the de-drying landslide influencing factors and displacement data into trend terms and periodic terms. For the trend term and periodic term, the second-order exponential smoothing (DBS) method and the dynamic multi-swarm particle swarm (DMS‑PSO) optimization extreme learning machine (ELM) model were used to predict the displacement. Finally, the predicted periodic term displacement and trend term displacement are added to obtain the total predicted landslide displacement. Through the periodic term displacement prediction model, the global optimal solution can be better solved, so that the prediction accuracy and reliability are higher.

Description

technical field [0001] The invention relates to the field of computer mathematics, in particular to a landslide displacement prediction method of a hybrid machine learning model. Background technique [0002] Landslide is a common geological disaster. It refers to the rock and soil on the slope, which is affected by internal and external factors such as groundwater, rainfall, earthquakes, and human industrial and daily activities. It is a natural phenomenon of sliding down the slope as a whole or dispersedly. The frequent occurrence of catastrophic accidents caused by landslides brings huge hidden dangers to the safety of human life and property. It is of great significance to strengthen related research on landslide prevention and control. The research on landslide displacement prediction is related to the site selection of many projects and the safety of life and property of surrounding residents; landslide prediction can win valuable for the safe transfer of people and p...

AI Technical Summary

Problems solved by technology

[0004] At present, the traditional landslide displacement prediction method adopts the prediction model to establish complex mechanical equations and statistical models according to the geological environment, which requires a large amount of calculation and a long modeling period. At the same time, with the development of technology, the data information obtained is increasing. Low data usage leads to a lot of data waste

Images