Research On Data Processing Method Of Deformation Monitoring

The main purpose of the deformation monitoring is through analyzing the deformation information of the deformable body to to make accurate predictions for its future trends as a result of which can provide a scientific basis and accurate guidance for the design of projects and buildings as well as the prevention and controlment of geological disasters. Whether we can basing on the requirements of projects to make scientific and rational analysis and modeling of the deformation information of the deformable body or not is of great significance to the whole deformation monitoring.Currently, the main analysis and forecasting model of deformation monitoring data are as follows: regression analysis model, gray system analysis model, time series analysis model, the Kalman filter model, artificial neural network model and spectrum analysis and so on.However, in real life, there are many reasons for the structural deformation and any kind of model can hardly accurately simulate and predict the deformation process of the structure. Today, there are a lot of scholars at home and abroad doing a lot of researches on the deformation monitoring portfolio forecast model.However, they do not form a complete theoretical system of it.In our survey work, the climate and environment, methods of observation, observation instruments and observers factors and many other reasons may lead to the loss or incompleteness of observation data failure and if taking unreasonable methods for processing it will greatly reduce the forecast accuracy of the deformation of the building and when the problem gets more serious,it will even make the whole deformation forecasting becomes meaningless.In the view of the above, this paper has mainly done the following pieces of work:1.The thesis analyses the modeling method of deformation, compares the accuracy of engineering building settlement prediction processing measured by several ways by experiments, and raises the scope and performance of these methods. 2.The thesis analyses the combination forecasting model applied in the analysis and prediction of deformation monitoring experiments, raises the main methods and basic theory of combination forecasting model, and investigates the problems should be noted in the application of combination forecasting model.3.The thesis analyses the common methods and basic theory of incompletdata processing, and summarizes the advantages, disadvanta ges and the range of application of these methods. It also apply al gorithm to AR(P) forecasting model when it deformation monitoring data is missing, and verifies the practicality and feasibility of theme thods by experiments.