摘要
底板突水预测是保证煤矿安全生产的重要环节,对煤矿的正常生产有着重要的意义。突水系数理论因其方便可操作性在底板突水预测中有着广泛的应用。本文通过实例分析指出突水系数理论在理论突水系数与临界突水系数相差较大时有较好的预测效果,但在临界突水系数附近却存在明显的波动性。为弥补突水系数理论的不足,本文引入BP神经网络对临界突水系数附近的样本进行预测分析,预测结果与实际情况相符,预测效果较好。建议在底板突水预测中首先采用突水系数理论进行预测,对其中处于临界突水系数附近的样本进行BP神经网络预测,可以保证预测效果。
Irruption forecast is the important link of guarantee safety production, to the normal produc- tion of mine has an important significance. Irru tion has a wide range of applications, this artic pt le ion of operability, due to its convenience in irruption predicby case analysis pointed out that the irruption of the coefficient in the theoretical and critical water gushing water coefficient widely better prediction of the effect, but in near critical water coefficient is significant fluctuations. To compensate for the water gushing shortage of, this article introduces the BP on critical water coefficient near samples for predictive analysis, forecasts and the actual situation in line with forecasts. Recommended in the prediction of inrush of water through water coefficient theory predicted, in a critical water coefficient near samples for BP neural network to forecast, you can ensure the prediction results.
出处
《山西焦煤科技》
2010年第9期36-39,46,共5页
Shanxi Coking Coal Science & Technology
关键词
底板突水
预测
突水系数理论
波动性
BP神经网络
Floor water invasion
Prediction
Water bursting coefficient
Volatility
BP neural network
