摘要
煤与瓦斯突出是地应力、瓦斯和煤岩力学性质综合作用的结果,其作用过程属于非线性问题,使用常规方法确突出预测敏感指标及其临界值,存在着效率不高的问题.采用三率法,研究了大平煤矿煤与瓦斯突出预测指标,得出了预测突出的敏感指标及临界值;通过使用人工神经网络的BP模型,选取样本、训练样本和识别样本,并与现场实际对比,对用常规方法所确定的敏感指标及其临界值进行了验证.结果表明,可以使用神经网络方法预测煤与瓦斯突出危险性和确定突出预测敏感指标.
Coal and gas outburst is a comprehensive result of earth stress, gas, coal and rock mechanical properties, which the process is a nonlinear problem. When using conventional methods for exploring forecast sensitive indicators and their threshold, the efficiency is not high. In Daping coal mine, forecast sensitive indicators and their threshold were obtained by using three percentage methods. By using the method of neural BP network which includes samples selecting, samples training and samples identifying, compared with the actual scene, it tests the results of sensitive index value determined by the conventional method. The results indicate that the method of neural network can be used in the prediction on coal and gas outburst, and determined sensitive index value.
出处
《河南理工大学学报（自然科学版）》
CAS
2008年第4期381-385,共5页
Journal of Henan Polytechnic University（Natural Science）
关键词
瓦斯突出
神经网络
敏感指标
gas outburst
neural network
sensitive index
