Research On Deformation Analysis And Prediction Model Of Grey BP Neural Network Based On Kalman Filtering

As an important means of rapid economic development, some large and very large projects continue to invest in construction, the construction process operations security issues are the focus of attention, to monitor their safety during deformation analysis and prediction.. Based on the way of deformation analysis and prediction can help to understand the rule of engineering construction and provide important decision basis for the safety evaluation and operation. At present, in the aspect of deformation analysis and prediction, there are a variety of deformation analysis and prediction models, including a large number of single model or combined model by single models such as the grey system theory, the BP neural network, regression analysis and so on. The paper elaborates that building the gray BP network based on Kalman filter after handling the observation data by Kalman filter and research on the gray system theory and BP neural network. The specific contents of the study are as follows:Part one. The part elaborates the development of deformation analysis and prediction at home and abroad, introducing the research status of the gray theory, Kalman filtering, BP neural network and the combined model. Then it summarizes the time series, the regression analysis method, the BP neural network and finite element method basic theory and characteristics.Part two. It elaborates the advantages and disadvantages of the grey model and the BP neural network in the single model condition, introduces the characteristics of the combination and several kinds of combination form. Additionally, elaborating the foundation of gray system and Kalman filter and the evaluation standard of each model, summarizing the basic form and the calculation process of the gray model based on Kalman filter and gray BP neural network.Part three. As the observed value is affected by the noise in this article, using the regression analysis, the BP neural network based on Kalman filtering, gray model and gray BP neural network based on Kalman filter to conduct the deformation analysis and forecast.Through comparing the model calculation result,it can get that the gray BP neural network based on Kalman filter has better accuracy level in this example.Part four. It points out some improvement and development prospects of the gray BP neural network based on Kalman filter. This article selected a factor which restricts the applicability of the model, the precision of the GM(1,1) model will have a possibility of further study.