Application Of Grey Markov Model In Settlement Prediction Of Ningbang Super High-rise Building

With the acceleration of urbanization,super high-rise buildings are springing up,and the safety problems of super high-rise buildings are increasingly prominent.In order to ensure the safe construction and use of the building,it is necessary to carry out long-term deformation monitoring of the building,and make accurate analysis and judgment on its deformation development trend,so as to provide reliable basis for the construction and operation of the building.Based on the settlement monitoring project of Ningbang super high-rise building and uses the actual settlement monitoring data as the modeling basic data,this paper analyzes and studies the settlement prediction of super high-rise building.The main research contents are as follows:(1)Due to the fact that the actual monitoring data may be non-equidistant and the single-point gray model cannot reflect the overall deformation trend and deformation law of the deformed body,the non-equidistant multi-point gray model is used to model and analyze the settlement monitoring data.(2)Aiming at the shortcomings of multi-point grey model,firstly,the weighted average method based on full information is adopted to optimize the initial values.Secondly,the combination of relative error and time distance is proposed to weight the original sequence.Finally,Simpson integral formula is adopted to optimize the background value and the optimized non-equidistant multi-point gray model is constructed.(3)Using the Markov model to modify the predicted value of the optimized non-equidistant multi-point gray model,an optimized non-equidistant multi-point gray Markov model is proposed.(4)Using Matlab to carry out program design,four kinds of deformation prediction models are established,namely: non-equidistant single-point grey model,non-equidistant multi-point grey model,optimized non-equidistant multi-point grey model and optimized non-equidistant multi-point grey Markov model.The reliability of the optimized models is compared and analyzed through engineering cases.Research and analysis show that,compared with other models,the use of optimized non-equidistant multi-point Markov to predict the settlement of super high-rise buildings can effectively improve the fitting and prediction accuracy.