The Application Of Combination Of Improved GM(1,N)model And Time Series Theory Model On Displacement Monitoring Of Dam

Displacement is an important indicator of the dam safety state, Analysis of damdisplacement monitoring data related to the safety of the dam and people’s lives andproperty within the dam area, therefore, strengthen the dam displacement monitoring isa great sense of work to protect the dam.Displacement monitoring of dam is usually affected by water pressure, temperature,aging and other factors, which are mutual restraint and interacting, and thus make thedisplacement monitoring of Dam an non-linear system. In order to analyze thedisplacement monitoring data of the dam accurately and do real-time forecasting, thispaper do some analysis and research on displacement monitoring data of dam by thecombination of improved GM(1,N) model and time series models, based on the featuresof displacement monitoring of dam and analysis of displacement monitoring data andmonitoring methods.Considering that the dam is affected by many factors and combining theapplication characteristics of the GM(1,N) model, this paper built GM(1,N) model to dosome analysis and research on displacement monitoring data of dam. The main work ofGM(1,N) model building is the selection of factors and compilation of the modelprocess, by selecting appropriate factor, GM(1,N) model can be built on displacementmonitoring data of dam. Accuracy of the model is a standard to judge the merits of themodel, in order to improve the fitting and forecasting accuracy when using GM(1,N)model, this paper improved the GM(1,N) model based on Simpson formula, and usedthis model in displacement monitoring of dam, and compared the fitting and forecastingeffect between traditional GM(1,N) and improved GM(1,N).For the fitting residuals of the improved GM(1,N) model, since the randomnessand dynamic of the fitting residuals, and time-series analysis is an effective method toprocess such data, therefore, this paper built ARMA model to process the fittingresiduals. Based on the analysis above, this paper did some analysis and research ondisplacement monitoring data of dam by the combination of improved GM(1,N) modeland time series models which is built from improved GM(1,N) model and ARMAmodel.This paper established an instance based on the displacement monitoring data of anarch dam, and did some comparative analysis of the model results by the analysis ofmonitoring data. Research had shown that the improved Simpson method used in this paper had some applicability and the improved method had achieved good results in theproject examples. Secondly, the combination model were better than the improvedGM(1,N) model in predictions. Therefore, the combination model can reflect thedisplacement of the dam more accurately, thus provided an effective tool for theapplication of displacement monitoring data of dam and the guarantee of the dam’s safeoperation.