A Multivariable Grey Model and its Application to Subgrade Settlement Prediction

Abstract

Subgrade settlement is a complex systematic process. Frequently used single point forecasting models can’t consider the correlation of settlement between the discrete monitoring points, so that it can’t represent the integrated deformation regularity of subgrade. For resolving this problem, this paper extends the single point GM(1,1) model into a multivariable grey model (MGM(1,n)), and completes the programming with MATLAB. Finally, A MGM(1,3) model is established to predict the subgrade settlement, and the result of the model is more precise and effective than the single point GM(1,1) model through a practical application.