Tunnel surrounding rock stability prediction using improved knn algorithm

Abstract

Accurate prediction of the stability of rock surrounding a tunnel is important in order to prevent from rock collapse or reduce the hazard to personnel and traffic caused by such incidents. In our study, a KNN algorithm based on grouped center vector is proposed, which reduces the complexity of calculation, thus improving the prediction performance of the algorithm. Then, the improved KNN algorithm was applied to the surrounding rock stability prediction of a high-speed railway tunnel, which, to our knowledge, forms the first application thereof for the prediction of surrounding rock stability. Extensive experimental results show that our proposed prediction model achieves high prediction performance in this regard. Finally, a laboratory experiment of a tunnel is conducted to evaluate whether the tunnel surrounding rock is stable or not. The experimental results matched the prediction results obtained by our proposed prediction model, which further demonstrates its effectiveness.