考虑钻孔速率的公路隧道围岩类别超前分类研究

为了有效的进行公路隧道围岩稳定性超前分类,将台车钻孔速率引入到围岩稳定性评价指标体系中,通过增加动量因子和设置自适应调整学习率的方法建立了改进型BP神经网络的围岩分类模型。该模型考虑了钻孔速率这一实测性、超前性的数据,为模型注入了更为丰富的信息量,提高了围岩分类的可靠性,能较准确的预测出掌子面前未开挖段的围岩类别,弥补了施工地质超前预报中靠主观经验观察、分析的不足。通过实例分析,所得的分类结果与实际吻合,对施工的指导性强。

Study on Surrounding Rock Stability Classification Considering Drilling Rock Rate Based on Artificial Neural Network

To improve highway tunnel's rock stability classification,the authors introduced the drilling rock rate as included in the classical surrounding rock stability appraisal system,developed a new classification method based on improved back-propagation ANN and considering drilling rock rate.And this ANN increases the momentum gene and adaptive variable step-size arithmetic.Drilling rock rate is considered as a measuring and preceding data and infused into the more abundant information for the model.Using this method,the surrounding rock of un-dug sector can be sorted out and predicted more accurately.It can be used as a site surrounding rock predication method for tunnel construction.