Study on Multivariate Selection Method of Rainfall Landslide Monitoring
In order to improve the accuracy of rainfall-type landslide monitoring,and solve the problems of false alarms that may be caused by relying only on rainfall monitoring as the basis for determining landslides in traditional rainfall-type landslide monitoring,a new multivariate selection method based on rainfall correlation is proposed to monitor landslide events.Firstly,the variables are screened by quantitative analysis of the correlation between the monitored physical quantities on site and the rainfall,then the principal component analysis method is used to reduce the dimensionality of the screening results,and finally the Gaussian mixture model is used for landslide monitoring.The results show that the multivariate selection method based on rainfall correlation can realize the multivariate selection of rainfall-type landslide monitoring and provide reasonable input variables for the landslide monitoring model.The research results can provide reference for rainfall-type landslide monitoring.
万方数据
维普
