Establishment And Application Of Prediction Model Of Foundation Pit Deformation Monitoring

The rapid development of urbanization, constantly improve the level of the national economy and people’s livelihood facilities for a growing variety of needs, and promote the rapid development of infrastructure. Construction of all kinds of high-rise facilities, subway, underground shopping malls and other facilities, bring the surge in the number of various excavation. Continue to refresh the record depth of excavation, the excavation area behind expanding, hides a lot of engineering safety problems, so pit deformation during the construction phase of the operation is essential. Binding pit deformation data acquisition, analysis of foundation pit deformation characteristics, as well as the use of certain methods of its deformation trend more accurate prediction can help engineers accurately grasp the construction and take preventive measures. Based on the above, for the excavation deformation monitoring and forecast research is necessary and significant.From the foundation pit deformation monitoring and forecasting of to foundation pit deformation monitoring project example as the background, using wavelet analysis tools and the gray forecast model and the monitoring data for processing and analysis, draw a conclusion:(1)According to the deformation of foundation pit monitoring examples, summarizes the foundation pit deformation monitoring technology of each focus, such as monitoring steps, buried, the monitoring data analysis, grasp the foundation deformation characteristics.(2)The wavelet analysis theory and method are introduced in this paper. The emphasis is on the wavelet threshold denoising method. Through the of different wavelet function, decomposition layers, the threshold value selection criteria, threshold function, weight transfer methods comparison, the use of signal to noise ratio, the mean root mean square error and smoothness as measure standard, completed the example data de-noising processing. Using the wavelet analysis method for monitoring data inspection found no singular value, singular value influence.(3)The stationary data of deformation in the foundation pits and wave deformation data to establish GM(1, 1) model to predict that the prediction accuracy for wave data needs to be improved. In accordance with the principle of new information, respectively, the establishment of the different dimensions of the original sequence was predicted, the level of accuracy and the establishment of sequence related to the dimension, but not the higher the dimension precision is high, and to for different data sequence is built to meet in the prediction of the optimum dimension.(4)After denoising the data to build wavelet grey combination model, with single prediction model were compared, and get the precision of combination prediction model is higher to verify the wavelet grey model in the foundation pit deformation monitoring and forecasting application feasibility.