| Landslides cause much damage annually to property and many people to run for their lives.The destroyed infrastructure and public transportation will take a lot of time and money to repair.Rainfall has been discovered to be the main source of the majority of landslides.An effective technique for avoiding casualties of landslides is the early warning of rainfallinduced slope instability.Since elastic wave signal travel can carry a string of information representing the physical characteristics inside the soil,it shows a new route to early warning.Early warning of rainfall-induced landslides based on elastic wave velocity could restrict a bit of lack of the conventional technique,including water content monitoring,slop movement monitoring,and rainfall monitoring,which can only monitor a bounded area with sensors themselves.Researchers have found that elastic wave velocity decreases with the volumetric water content and accelerated decreases when the slope moves,so they recommend giving an early warning at the point of wave velocity accelerated decreases.However,there is much noise in the elastic wave signal collected by bender elements type piezoelectric transducers.Travel time in P wave or S wave does not have precise positioning because of used conventional denoising(Superposition Denoising)that can’t eliminate efficiently highfrequency noise.In addition,calculating elastic wave velocity need a high-frequency data acquisition equipment,and it is constrained by the observational method.So,it is needed now to optimize an elastic wave denoising technique,and give an early warning project based on elastic wave further.This study aimed to study early warning mechanism of rainfall-induced landslides based on elastic wave amplitude.One provided an elastic wave denoising technique,combining the wavelet transform and the superposition method,which means the elastic wave signals(collected by the bender elements type piezoelectric transducers)were superimposed on each other before proceeding with the wavelet threshold denoising.To verify the applicability of the method,the signal simulation tests were used to compare four different signal-denoising methods.Another recommended using elastic wave amplitude to monitor soil slope,which reflects soil moisture and deformation and warns when the amplitude reaches the threshold.The elastic wave amplitude response characteristics in the soil moisture and deformation are mutually verified from static and dynamic viewpoints by the model box test and the slope model test,respectively.The signal simulation tests show that the combined superposition-wavelet threshold denoising method optimized the elastic wave signal processing in landslide warning with elastic waves,increased the accuracy of the elastic wave signal,and improved the efficiency of the elastic wave generator.The model box tests show that the elastic wave amplitude decreased nonlinearly with increasing volumetric water content in the soil of each dry density.The slope model tests show that the deformation could cause an increase in the wave amplitude,and its change range can reflect a collapse range in the slope.The elastic wave amplitude data after eliminating the trend term by empirical modal decomposition could more intuitively describe the relative change in wave amplitude,the relative change in wave amplitude.The relative change wave amplitude is greater than 0,which means the slope has deformation,and the warning can be issued at this time to start the risk avoidance work efficiently. |
