Study For Dam Deformation Monitoring Data Based On Time Series Analysis

Dam safety monitoring plays a crucial role in ensuring dam safety operation and people’slives and property safety. Dam deformation monitoring is an important aspect in the field ofdam safety monitoring, it usually establishes mathematical model to monitor and judge thesafety situation of the dam through the dam deformation observation data analysis.As forsome commom model such as statistical model and deterministic model, they all need factorsin modeling, and sometimes it’s difficult to modle some dam deformations use a certainfunctional relationship, or factor series often miss information.Time series analysis has the ability to analyze and forecast the trends of complex systems,the core idea is to predict future information using their own historical information andconsidering various factors combined effects. Under such situation, this article will do someresearch on dam and some other hydraulic structures deformation monitoring data using timeseries analysis,The main research contents and results are as follows:(1) Discussed and summarized the steps of establishing dam deformation statisticalmodel, and put forward an idea to find a best statistical model by selecting best combinationof factors and model. This thesis is based on Changzhou Hydro-complex which is located inGuangxi province. In this paper, the quantitative relationship between the ship-lock of theproject settlement and deformation factors was established by selecting best modle incombinations of a group of factors and two statistical models.(2) Using time series analysis to establish the ship-lock control building settlement pointsARIMA model, which got a better modeling effects and higher prediction accuracy,resultsshow that the subsidence of the area has been basically stabilized with no significant growingtrend.It found that when the trend of the series is obvious,the seasonal nature is easily coveredby its trend,ARIMA model is not easy to discern its seasonal nature.(3) Separating the time series into trend part, cycle part and its residuals, modlingresiduals by time series analysis, then established the cycle-ARMA exponent compositemodel which was used in ship-lock control building point ZLD-02, its results has littledifferenece with ARIMA modle.(4) Modling and forecasting some sluice gate and ship-lock typical points in this example project using the seasonal cycle-ARMA composite model, results show thatsubsidence points revealed obvious seasonal features with12months cycle interval and noobvious trend, also individual points with short cycle period. It can be concluded that theseparts of the settlement is caused by the cyclical temperature changes.(5) The article also set up a deformation significant factors multidimensional CARtiming model based on external river navigation lock points WLD-08, it was found that theeffectiveness and accuracy of the model has no much difference with the seasonalcycle-ARMA composite model, indicating that this method has a certain prospect inforecasting dam deformation.(6) Compared timing model (the seasonal cycle-ARMA composite model,multidimensional CAR model) and statistical models (multiple regression, stepwiseregression)，the former is much better than the later and the seasonal cycle-ARMA mode isjust lightly better than the multidimensional CAR model, but the reasons of deformationcouldn’t be given by the composite model, while the CAR model could gave quantitativeexpression of significant factors, both have pros and cons.Time series model has good consequence in analysis and forecasting dam deformationwith higher prediction accuracy in predicting the deformation data, especially for short-termforecasts.At the same time, the time series model has dynamic epitaxy, updating thetime-series data, it will be able to maintain a high level of short-term forecast accuracy. Inshort, the application of time series analysis method can accurately reflects the discipline ofdam deformation, forecasts with high accuracy, makes early warning to the deformation ofdams and other large hydraulic structures as soon as possible, then achieves the goal ofmonitoring the safety of the works and ensuring the building safety operation.