Research On Cloud Analysis System Of Deformation Monitoring Data Based On Kalman Filter

The cloud analysis service of engineering deformation monitoring data is one significant development direction of surveying and mapping informatization.The rapid development of commercial cloud computing has laid foundation for the technology framework of above system.Cloud analysis of monitoring data must be based on the high-efficient management of distributed massive deformation monitoring data.Large project of high-speed railway foundation deformation monitoring contains wide lines and many sections.Standardization and informatization of monitoring data form these projects is a challenge for companies;Multi-sensor integrated development of traditional engineering deformation monitoring solution makes data source types for single project increase.However,it's difficult for the current localized or singularized management system to satisfy multi-project integration and network management needs from users.In order to solve above problems,this paper constructs the solution of management,cloud analysis and early warning for the deformation monitoring data based on B/S architecture.The key technology of each link has obtained the following research results:1.Through studying new requirements of monitoring data management and technical architecture of B/S cloud system,this paper constructs the solution of cloud monitoring system for engineering monitoring based on the technical process of "early warning value extraction-cloud analysis and forecasting-early warning information release".2.This paper designs a distributed form storage structure which can optimize the retrieval efficiency of massive deformation monitoring data,and the extraction of early warning information from multi-source monitoring data is realized based on the database stored procedure mechanism.3.For the problem that the mathematical model and priori noise of engineering deformation system are inaccurate during the process of online cloud analysis based on Kalman filter,this paper studies modeling process of different state model Kalman filter as well as adaptive Kalman filter based on maximum likelihood criterion,which provide reference for the comprehensive discrimination and network release of early warning information.4.Based on the simulation data and the measured data,this paper demonstrates the improvement effect of adaptive Kalman filter;In order to solve the problem that large disturbance of process noise caused by inaccuracy innovation sequence,an improved Scale-Q adaptive filter algorithm based on the innovation variance adjustment factor is proposed in this paper,and the results show that the improved model has better noise reduction effect.Through the above solution,this paper realizes the unification of the data management subsystem for the high speed railway foundation deformation monitoring and the centralized management subsystem for the multi-source deformation monitoring data under the B/S architecture,which expands the application of B/S architecture deformation monitoring system.