Research On Missing Data Filling Algorithm For Automatic Monitoring Data Of Slope

In today's society,there is a widespread phenomenon of data deletion,such as experimental research field,statistical research field and engineering field.The lack of data means information incomplete,means that increases the difficulty of data processing and analysis,reduce greatly the accuracy of the data analysis results,reduce the working efficiency of staff,have the effect of the wasted effort.In the slope in the process of the automatic monitoring,data acquisition and transmission are generally use of various types of sensors and other electronic equipment,automatic monitoring system in the outdoor environment,inevitably appear mostly wear,aging,lack of power wait for a phenomenon,these can all result in a lack of monitoring data,leading to slope stability evaluation appeared deviation.Based on 2016 ~ 2017 in ShaZhuang zengcheng guangzhou to huadu Beixing street phase ii project(li city to huadu Beixing section)project JK02 somewhere in the contract period of high cutting slope safety monitoring automation data,for example,in view of the lack of automatic monitoring data to fill the content,the main work is as follows:(1)give a detailed introduction to the research background,significance of the paper and the research status at home and abroad;This paper analyzes the common cause of data loss in various social research fields,and explains the classification method of missing data and the types of missing data.(2)research on missing data processing methods in detail,this paper expounds the rule of filling missing data,and introduces in detail the two missing data filling algorithm,namely the Elman neural network algorithm and time sequence algorithm.(3)according to the characteristics of the automatic monitoring of slope,the lack of analysis of the slope in the process of the automatic monitoring data,introduced several common data types,missing data with the same characteristics of these types of missing is the lack of a certain period of time;Slope surface slope monitoring data,for example,to verify the correlation between different time intervals of the same data,it is concluded that the missing data has the characteristics of time series,and finally by comparing the Elman neural network algorithm and time sequence forecasting model for different data missing review probability to fill the data experiment,through the absolute error and root mean square error to judge its filling effect,review concluded that time series forecasting model to deal with this type of data,missing data filling effect is better.Missing data fill in today's era of big data is of great significance for data mining,and with the progress of computer science and technology,more and more algorithms have been proposed to fill the missing data,this research will also be more deeply.The purpose of this thesis is to provide some reference value for the missing data processing.