Study For Foundation Pit Settlement Monitoring Data Based On Time Series Analysis

With the accelerating urbanization in our country, the number of tall and super-tall building(structure) grows rapidly. Excavation of foundation pit engineering settlement plays a crucial role among the affected factors of normal use and security of buildings, it is indispensible to monitor the foundation pit engineering systematically. Meanwhile, time series analysis model has good results in dealing with dynamic data and has been widely employed in various industries. Applying time series analysis model to deformation monitoring, building a mathematical model through long-term accumulation of monitoring data to understand and grasp the variation of the development of things. Giving an effective analysis and forecast of the future state of the foundation and building at the same time. This paper is about to do some exploring research in this area.The basic thought and method of time series will be expounded as well as several common time series models and the dynamic characteristic of time series. The model will be predicted by using the autocorrelation function and partial autocorrelation function in time series analysis. The comparison among several kinds of order code predicated in time series will lead to the best order criterion used in building model and forecasting the deformation. A detailed introduction of building time series analysis model, based on the deformation monitoring data of foundation pit monitoring of the basement engineering of Xinyihuating in Nansha district, Guangzhou city, will be given and used to analyze the forecast results at length.Due to the frequent observations, large amount of observation data and complex model classification, the theory of least squares estimating methods will be employed to improve the time series. The establishment method and the application of variance-covariance function of the least squares estimating method will be clarified. The prediction is calculated by using the same sample data, which will be comparatively analyzed with the prediction results of time series analysis. Preliminary evidence shows that the least squares estimation can be successfully applied in foundation pit engineering settlement forecast.