Deformation Monitoring Of Surrounding Buildings During The Period Of Foundation Pit Construction And The Research Of Forecasting Methods

In recent years, with Chinese rapid economic development, many large foundationpits which exist in the cities or residential areas have appeared. The construction of thefoundation pits has brought some safety hazard to the surrounding buildings. It is animportant mean, which are real-time deformation monitoring of the building, analysisand processing of measured data and prediction of deformation, to ensure the safety ofthe surrounding buildings during the foundation pit construction.This paper briefly introduces the basic concepts of deformation monitoringtechnology, then makes related discussion on the layout of monitoring controlnetwork, the layout of reference point and work point, monitoring methods, dataprocessing, deformation analysis, etc, with an instance of “Surrounding buildingsDeformation monitoring project attached to Bai Fu Lu residential district shantytowns Reconstruction Pit Engineering”. Starting from aspect of theoretical analysis,this paper has introduced the basic theory of regression analysis, gray system andneural network，then analyzes a variety of factors affecting the deformation ofbuildings and establishes accurate forecasting models combined with the instance of“’Bai Fu Lu’ Monitoring Project Engineering”; Because the calculation of deformationforecasting model is very complicated, the author uses MATLAB software toimplement the computer program of calculation of the deformation model analysis.This paper uses the unified measured data to analyze comparative effect of theforecasting models, then summarizes the merits and applicability of the singleforecasting models and the combined forecasting models in engineeringapplications; The results show that the the neural network model and the gray neural network model, the forecasting effect of which is best, could fit the data sequencemost accurately, because they not only could analyze the multiple impact factors butalso could find the potential deformation laws of the data sequence; Coming toconclusion, for the prediction of surrounding buildings deformation caused by thefoundation pit excavation, the neural network model and the gray neural networkmodel, which have strong practicality and reliability, are able to characterize thedeformation information and laws.Through the practice, this paper summarizes some useful conclusions ondeformation monitoring and forecasting methods of surrounding buildings during thefoundation pit construction, which has a certain practical significance and is able toprovide a reference for similar projects in the future. Finally, the author brieflyillustrates the deficiencies and research to be done about this article.