The Comparison Of Pretreatment Methods And The Study Of Modeling About Slope Monitoring Data

The safety of slope engineering was related to the property safety and people’s lives.slope safety monitoring is the direct indicator which guiding the construction, operation,production and maintenance of slope, it plays an increasingly important rule on predictingworking condition of slope, feedback slope monitoring information and ensuring slopesecurity. The slope displacement is the most important monitoring physical variable inslope safety monitoring which can reflect the slope deformation directly. As to slopedisplacement monitoring, fixed inclinometer is given a widespread attention in slopeengineering for continuous monitoring and automatic acquisition and remote operation.Analysis and feedback of monitoring data is essential to the slope safety monitoring. Itis inevitable that there is a large number of non-real fluctuations and mutations and the lackof monitoring information in different degree in the actual engineering, which will bring anegative impact on post-analysis of monitoring information. Considering this, in order toeliminate the non-real fluctuations and mutations of monitoring data effectively, this papercombine with the actual engineering takes slope displacements obtained by fixedinclinometer as the object of analysis, then uses the window moved polynomial smoothmethod and wavelet to remove the noise of monitoring data, Depending on the differenttype of noise to explore the use of appropriate de-noising method to extract the slopedisplacement trends and to obtain more clean slope displacement monitoring data.As to the lack of monitoring data, in order to maximize the excavation data andimprove monitoring forecasting effect, this paper combine with the actual engineering andinterpolation work experience to take Lagrange interpolation as the main method ofinterpolation analysis method for exploring the selection and application in different formof data missing. The results show that monitoring data obtained by the reasonableinterpolation get a complete set of monitoring data, meanwhile, it also reflects the actualtrend of the entire data segment.The complex process of slope deformation is deemed to be gray process in a certainrange and a certain time zone, take slope displacement as gray variable changed in a certainrange. This paper takes Gray theory as the theoretical basis, Gray forecasting model andassessment model as its primary research tool, the slope displacement data as the object ofstudy for exploring the impact on modeling effect of the post-monitoring data frommonitoring pretreatment work. The results of modeling analysis of the originaldisplacement monitoring data and de-noising monitoring data show that de-noisingpreprocessing effectively reduce interference on monitoring data from noise informationand increase this mathematical model effect, then accuracy and reliability of model isimproved. Respectively analyze the modeling effect of monitoring data which includemissing and complete data after interpolation, then according to analysis and comparison ofeffect of fit and prediction to discuss the feasibility and stability of establish of the metabolic model based on interpolation results. Afterward, study slope deformation lawthrough two angles of time change and its spatial distribution based on clean and completemonitoring data obtain by pretreated work. On time changes, this take time function asrelevant factor and relevance as the basis of selection of time function for establishment ofa multi-variable gray model, then test the stability of their metabolism model. On spatialdistribution, according to gray correlation characteristics of measured point displacement atdifferent depths in the slope,this paper establish the gray nonlinear model for the spatialdistribution cumulative displacement value, the spatial distribution law of the cumulativedisplacement is obtained. finally, based on the modeling results, preliminary evaluation ofthe status of each measuring point based on Triangle Whiten Function gray assessment isestablished.