Font Size: a A A

Study Of Landslide Deformation Prediction And Early Warning Mechanism Of Synchronization

Posted on:2013-02-15Degree:MasterType:Thesis
Country:ChinaCandidate:F L CengFull Text:PDF
GTID:2230330395963244Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
As the world second largest geological hazards which is just after the earthquake, landslides cause great harm annually to the human production and life on a global scale, and it has the feather that occurred suddenly, cause large harm and governed difficultly. The present research tools include the methods through the landslide’s physical mechanism and its external environmental conditions while comprehensively consider the external and internal influence to predict landslide’s evolution trends. Confined by the complexity of landslide’s composition and influence factors, the physical study method still not formed effective model to describe the landslide’s evolution progress. In recent years, with the continuous development of landslide monitoring means and the introduction of intelligent algorithm theory, predictive analytics about the landslide evolution based on the landslide monitoring data has get a bigger development. This paper is a study about landslide evolution based on analysis of deformation monitoring data; Taking into account the highly nonlinear characteristics of the landslide deformation process, it introduces the empirical mode decomposition algorithm to make the landslide nonlinear signal analysis; Empirical mode decomposition algorithm is a method developed in recent years and has a wider application in nonlinear signal analysis, so it is introduced into this paper for the signal analysis of landslide deformation.First, displacement is one of the important information feedbacks from landslide deformation process, and when the medium complexity and physical, mechanical mechanism of landslide is rather ambiguous circumstances, it is an easy method to analyze the landslide with its deformation data. With the gradual development of the statistical mathematical methods, more and more artificial intelligence methods (neural networks, support vector machines, swarm intelligence algorithm) was introduced into the field of engineering analysis.Started from the analysis of landslide deformation monitoring data, this paper made a landslide deformation process prediction with the combination of EMD (Empirical Mode Decomposition) signal decomposition and support vector machine prediction algorithm, and then build a comprehensive landslide deformation prediction model to make a reasonable prediction about the landslide future trends. With the phase synchronization analysis about landslide deformation and variation of rainfall, it come out the results that there is not only existing space synchronization phenomenon between the monitoring point in the landslide process of change, but also time synchronizations between landslide deformation process and rainfall factors; This synchronization can provide better research ideas for future landslide prediction warning.
Keywords/Search Tags:landslide, deformation prediction, PSO-SVM, empirical modedecomposition, phase synchronization analysis
PDF Full Text Request
Related items