基于时变参数的灰色模型在变形监测中的应用OACSTPCD

Aiming at the low fitting and prediction accuracy of traditional GM(1,1)prediction model,we established a grey model based on time-varying parameter in this paper.Firstly,we introduced the establishment processes of traditional grey GM(1,1)and second-order GM(2,1)model.Then,based on the principle of time-varying parameters to optimize grey action,we established grey GM(1,1)and GM(2,1)model based on time-varying parameter.Finally,taking the monitoring data of surface subsidence in the mining area for example,we carried out the prediction comparison and the accuracy evaluation of gray model through different gray model schemes.The experimental results show that GM(2,1)model based on time-varying parameter has the highest prediction accuracy,and the gray model accuracy test level is level Ⅱ,illustrating GM(2,1)mod-el based on time-varying parameter is more suitable for the deformation trend of surface monitoring data in this mining area.