weighted Markov landslide early warning method based on system clustering

Abstract

The invention discloses a weighted Markov landslide early warning method based on system clustering. The method comprises the following steps: collecting a displacement speed data sequence of a certain monitoring point; dividing the n initial sample speed data sequences into n equal parts; calculating the distance between the speed data of the initial node by adopting a gravity center method; drawing a displacement speed state system clustering tree; determining an initial probability vector of the model; solving a model autocorrelation coefficient; obtaining a probability distribution vectorof the future state of the m-order weighted Markov prediction; judging whether landslide early warning should be carried out or not; updating the data; and landslide early warning effect analysis andsample capacity optimization. The landslide early warning method starts from reasonable selection and description of landslide criteria, gives consideration to timeliness, interference resistance andcredibility of early warning, explores a new landslide early warning method, and provides a scientific means for accurate and timely implementation of landslide early warning.

Description

technical field [0001] The invention belongs to the technical field of open-pit mining, in particular to a weighted Markov landslide early warning method based on system clustering. Background technique [0002] With the continuous increase of the number and production scale of open-pit mines in my country, landslide disasters in dump sites occur frequently. Due to the spatio-temporal variation characteristics of geotechnical material properties, uncertainty and the complexity of factors affecting slope stability, the issue of landslide early warning has not yet been solved. Get well resolved. Most scholars use relatively conservative and empirically based thresholds of displacement rates and creep curve tangent angles as the basis for early warning, which often results in high early warning error rates and inaccurate warning areas. A small number of scholars proposed to use the displacement acceleration threshold as the landslide criterion for early warning, but they could ...

AI Technical Summary

Problems solved by technology

[0002] With the continuous increase of the number and production scale of open-pit mines in my country, landslide disasters in dump sites occur frequently. Due to the spatio-temporal variation characteristics of geotechnical material properties, uncertainty and the complexity of factors affecting slope stability, the issue of landslide early warning has not yet been solved. get well resolved

Most scholars use relatively conservative and empirically based thresholds of displacement rates and creep curve tangent angles as the basis for early warnings, which often result in high early warning error rates and inaccurate warning areas.

A small number of scholars proposed to use the displacement acceleration threshold as the landslide criterion for early warning, but they could not find a unified acceleration threshold to apply to different types of slopes with ever-changing natural and engineering conditions; Wang Dong et al. The criterion is described from the perspective of the slope, and a dynamic identification method for the moment when the slope is about to slide is proposed, which overcomes the lack of universality of the displacement rate threshold early warning in the past, and provides a new idea for the landslide early warning. However, due to the This method generally can only make a judgment after the start of the landslide, and there is a slight delay in the early warning

[0003] Therefore, the existing landslide early warning theories and methods still have problems such as low reliability, high error rate, or untimely early warning.

Images