According to the characteristics of roadway within soft rock, the factors that influence the stability of roadway within soft rock are analyzed. The optimized neural network theory is established about the soft rock roadway support program in order to provide a rational support program. The model has been applied well in an iron mine. Based on the nonlinear relationships between rock deformation and its influencing factors, the neural network model on the prediction of roadway deformation is built. Through comparing model analysis results with the actual results, the rational support form is put forward. All these researches provide the scientific rules for the support design and production management.
With the increase of the exploiting of the surface mining engineering, the complex exploiting of the soft rock mining engineering is also growing (Zhu & Shi 1997). Under the action of high stress, the characteristic of the surrounding rocks of roadway is soft, and the surrounding rocks begin to appear significant deformation. Many forms of the supporting in the soft rock roadway are not consistent with the deformation characteristics of surrounding rock, and the unreasonable roadway support parameter selection would lead to the poor effect of the roadway support, and consequently, the maintenance difficulties would seriously affect production and safety of the mine (Xia et al. 2003, Kuang & Song 2005).
Since there are many factors which would influence the supporting design of the soft rock roadway, and it belongs to the nonlinear problem, the traditional mechanical and mathematical methods can not solve these problems accurately (Yang & Wang 2006). With the ability of high level of self learning, the adaptability and nonlinear mapping, the BP neural network can be widely used in the optimization of the roadway supporting plan (Zhu et al. 2005, Gao 2003, Wei et al. 2010).
In this paper, based on the analysis of the influencing factors on the deformation of soft rock roadway, the neural network theory is applied to the optimization of the supporting plan and the prediction of the surrounding rock deformation.