基于神经网络的岩爆分级预测

Classification Prediction of Rock Burst Based on Neural Network

为了检验神经网络在岩爆预测分级中的可行性,利用神经网络的模式识别特性,构建了岩爆分级预测的神经网络模型。以13组实测数据作为神经网络的训练样本、5组实测数据作为测试样本验证神经网络的工作性能。验证结果表明,神经网络的岩爆分级预测结果与实际结果完全一致,表明训练良好的神经网络可以用于工程实践中的岩爆分级预测。

In order to test the feasibility of neural network in rock burst prediction and classification,a neural network model for rock burst classification prediction is constructed by using the pattern recognition characteristics of neural network.Taking 13 groups of measured data as training samples and 5 groups of measured data as test samples to verify the working performance of neural network.The verification results show that the rock burst classification prediction results of neural network are completely consistent with the actual results,indicating that the well-trained neural network can be used for rock burst classification prediction in engineering practice.