基于RES理论的岩爆倾向性预测方法

The method to predict rockbursts proneness based on RES theory

从工程地质因素、复杂环境因素和人为开挖因素3个方面分析了岩爆启动的主要影响因素,在此基础上,提出了一种基于RES理论的岩爆智能预测模型,并论证了人工神经网络的参数分析原理。此外,采用改进的前馈神经网络BP算法对交互作用矩阵进行编码以及对参数的相对交互作用强度进行了分析。研究结果表明:运用该岩爆智能预测模型,不仅使岩爆倾向性的预测具有动态特性,同时又可以方便地对岩爆启动的主控因素进行分析。

This paper analyses the main factors inducing rockbursts from rock engineering geology, environments and excavation technology. A new theoretical modeling approach to predict rockburst proneness and deduced the mechanism of parameters analysis of artificial neural network. An improved feedforward BP arithmetic is applied to coding interaction matrix and analyzing relative strength effect of theoretical modeling. The results show that the parameters analysis of artificial neural network can be used to dynamically predict rockbursts proneness, moreover, it can be used to analyze the strength of main factors inducing rockbursts.