Analysis Of Sites Stability And Sedimentation Prediction In City Subway Security Monitoring

City subway is an effective measure in the process of urban modernization and development to relieve traffic pressure. City subway in general is built in the downtown where the geologic conditions are complex, underground pipelines are in staggered arrangement, buildings are everywhere and population is dense. In addition, non-metro construction projects are becoming more and more along or near the top of the subway tunnel, the above-situations inevitably lead to a certain degree of sedimentation of subway tunnel. When the sedimentation amount reaches a certain limit, it will bring hidden troubles to subway constructions and normal operation. Therefore, real-time sedimentation monitoring of subway is of great significance. The analysis and processing of sedimentation monitoring data using computer technology can help the monitoring staff complete interior work scientifically and efficiently. The main research contents of subway construction monitoring are followed:1) According to the characteristics of subway sedimentation monitoring, it is discussed about the benchmark network layout method, main technical parameters and the principle of monitoring network in this paper. According to the characteristics of monitoring network, the right adjustment datum will be chosen after collecting monitoring network data. Monitoring datum network adjustment datum and its selection were analysed with an example.2) The main purpose of stability analysis of subway sedimentation monitoring datum network is to determine whether the monitoring datum net is stable or not. Several methods of stability analysis including the mean spacing method, tolerance test method and t-test at the same adjustment datum were introduced. Integrity test and single point test of the monitoring network were carried out by using the above methods with an example. 3) Gray system theory and building model, regression algorithm and the algorithm of support vector machine were discussed. Gray prediction model was combined with support vector machine. Firstly, sedimentation data add up to the new sequence with monotone increasing by the model of Gray system to enhance its regularity. Secondly, some sedimentation data with the way of interval rolling treated as input value were trained and predicted by support vector machine. Finally, the predictive values were made inverse accumulated operation. It is the procedure of GSVM combination forecasting model. The cumulative sedimentation amount of a certain point in sedimentation monitoring network were predicted by the method of combination forecasting model with an example in a certain city.4) By the Visual Studio2010platform and C#language, the engineering measurement data processing software suitable for city subway security and stability analysis and sedimentation prediction was developed.