Application of Metabolic GM(1,1) Model and Verhulst Model in Forecasting of Soft Rock Roadway Convergence

Deformation monitoring of soft rock roadway plays an important role in process of roadway construction, which is also a important method of estimating stability of soft rock roadway and grasping soft roadway deformation law. Traditional forecast method is general numerical calculation which is based on geotechnical mechanics and statistical fitting method, or other method. In this paper, according to soft rock roadway deformation data, programming is carried on which is based on MATLAB7.0 software; Metabolic GM(1,1) and Verhulst forecasting model of soft rock roadway deformation is established, and the deformation of roadway is forecasted. Through calculation, it is found that precision of complete data Verhulst model is higher than that of complete data GM(1,1) model, however, the precision of metabolic GM(1,1) is higher than that of metabolic Verhulst model. The reason is that, as soft rock roadway tends to be stable, the data of roadway deformation tends to be s-shaped, and insufficient data sequence could not reflect the variation characteristic of it. In this paper, calculation result has a certain practicability.