论文部分内容阅读
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。
There are many factors that affect the coal and gas outburst. When using neural network to predict, it is the key to select the salient predictors. The methods of choice based on experience and the so-called principle of “more, better, more” have some irrationality and unscientificness. The author uses the gray relational analysis to screen out the prominent forecast index and make the prominent forecast with the neural network modeling so that the selection of the outstanding forecast index can be converted from the qualitative analysis to the quantitative analysis and the combination of gray theory and neural network in the field of coal and gas outburst prediction . After example verification, this method is feasible.