基于灰色理论与神经网络的煤与瓦斯突出预测

Prediction of Coal and Gas Outburst Based on Neual Network and Grey Theory

影响煤与瓦斯突出的因素众多,应用神经网络进行预测时,选取突出预测指标是关键。基于经验和所谓"多多益善"原则的选择方法都有一定的不合理、不科学性。笔者应用灰色关联分析筛选突出预测指标,结合神经网络建模进行突出预测,使突出预测指标的选择由定性分析转化为定量分析,实现了灰色理论同神经网络在煤与瓦斯突出预测领域内的结合。经过实例验证,本方法是可行的。

The key to predict coal and gas outburst affected by many factors by neural network is to choose pa- rameters. It is unreasonable and unscientific in some extend to choose parameters either by experience or follow the principle so-called ＇ the more the better＇. The authors filtrate parameters by grey relevancy analysis and combine neural network model to predict. Qualitative analysis is transformed into quantitative analysis in terms of the choice of parameters. The actual example demonstrates that this method is feasible.