A new rock burst damage scale index (RDSI) based on the Rock Engineering Systems paradigm and on a BP-ANN is proposed. Seven parameters that are likely to affect rock burst damage are considered to construct the ANN and to objectively code the interaction matrix. A probabilistic analysis is employed to associate damage categories to RDSI values. The methodology is validated using 24 case histories that had not been employed during the training of network. Results suggest that the RDSI can probabilistically predict the damage scale with an accuracy that could be acceptable in engineering, hence simplifying the prediction for real applications.

中文翻译：

---

使用系统方法评估地下矿山的短期岩爆破坏

提出了一种基于岩石工程系统范式和BP-ANN的新岩爆破坏尺度指数（RDSI）。考虑了可能影响岩爆破坏的七个参数，以构建人工神经网络并客观地编码相互作用矩阵。使用概率分析将损坏类别与RDSI值关联。使用网络培训期间未使用的24个案例历史来验证该方法。结果表明，RDSI可以以工程上可以接受的准确度概率地预测损坏程度，从而简化了实际应用中的预测。