| With the development of new-type urbanization, the carrying capacity of city in resources and environment become saturated, the increasing population aggravate traffic congestion and environmental pollution, the quality of life is severely deteriorated,especially puzzled normal travel. To solve the problem we have to develop the underground rail, it bring a opportunity to develop the monitoring of underground rail. a variety of monitoring equipment and technology appeared, but the current technology still can not replace the manual measurement. The deformation monitoring of underground rail require the monitoring date exact and timely, with close connections to time, but with the influenced by the construction environment, the monitoring data has error, omission and so on. These problems are deeply troubling the monitors.The research of this paper is in order to solve this situation and provide methods to solve the problem for the monitors. Deformation monitoring data processing is mainly divided into two parts: data preprocessing and data analysis of the forecast period.Preprocessing stage includes the data of error identification and interpolation of missing data, introduces the corresponding theory at the same time, using the measured data to contrast and analysis which can improve the defects of some algorithms, make higher accuracy and reliability. In the data analysis and modeling phase which include the inflection point monitoring data, time series model test, modeling process, etc. According to the characteristics of the data, using the differential treatment of ARIMA model, in the process of modeling, use a combination of images and data form to makes the modeling process more simple and intuitive.The experimental date of this paper come from Qingdao no.2 engineering deformation monitoring’ s measured data, the data is reliable, this to other monitoring project in related fields, has reference significance. |
PDF Full Text Request