为了实现对瓦斯涌出量准确预测,从而有效预防瓦斯灾害.提出1种结合卷积神经网络(convolutional neural network,CNN)和双向长短期记忆网络(bidirectional long short-term memory,BiLSTM)的瓦斯涌出量预测模型,采用CNN在时间序列上提取瓦斯涌出量及其影响因素的局部关键特征,有效捕捉数据的局部时序相关性;BiLSTM模型利用这些特征,通过其前向和后向处理能力,全面捕捉时间序列中长期依赖性和复杂模式.研究结果表明:该模型预测准确率达93.6％,均方误差显著低于CNN、BPNN、LSTM、BiLSTM、CNN_LSTM、CNN_BiLSTM 6个模型,决定系数接近1,表明其出色的预测能力和解释力.研究结果可有效预测瓦斯涌出量波动,有助于提高矿井瓦斯风险预警能力,提升矿井安全管理水平.
Prediction model of mine gas emission based on CNN_BiLSTM
In order to achieve the accurate prediction of gas emission quantity and effectively prevent gas disaster,a predic-tion model of gas emission quantity combining the convolutional neural network ( CNN ) and bidirectional long short-term memory ( BiLSTM) network was proposed.The gas emission quantity and key local features of its influencing factors were ex-tracted from time series by using CNN,and the local temporal correlation of data was effectively captured.These features were used by the BiLSTM model,which leveraged its forward and backward processing capabilities to comprehensively capture the long-term dependency and complex patterns in the time series.The results show that the model has an accuracy rate of 93.6％,with a significantly lower mean square error than the CNN,back propagation neural network (BPNN),long short-term memory ( LSTM),BiLSTM,convolutional neural network-long short-term memory ( CNN_LSTM),and CNN_BiLSTM models.The coefficient of determination is close to 1,indicating its excellent predictive ability and explanatory power.The research re-sults can effectively predict the fluctuation of gas emission quantity,contribute to improve the gas risk warning capability,and enhance the safety management level of mines.
prediction model of gas emission quantityconvolutional neural networkbidirectional short-duration memory networkreverse neural networkbaseline comparison
万方数据
维普
