基于径向基的瓦斯涌出量灰色预测模型

为了进一步预防煤层瓦斯突出,实现准确、快速预测煤矿瓦斯涌出量的大小,首先采用1-AGO对样本数据进行处理,建立灰色(GM)预测模型,再利用径向基(RBF)神经网络对灰色预测模型结果进行预测,以作为其最终的预测值;采用阜新煤矿某工作面瓦斯涌出量的历史数据进行建模,实验结果表明,GM-RBF组合模型在预测精度及训练误差方面均优于单一的GM模型和RBF神经网络预测模型;算法计算简便,减弱了数据的随机性及模型误差,煤矿瓦斯涌出量的预测平均误差减小到1.57%。

Grey Forecast Model of Gas Emission Based on Radial Basis Function

In order to prevent coal seam gas outburst,to realize accurate,fast forecasting gas emission from the coal mine,first Processing data by 1-AGO,establish forecasting model,by using grey forecasting method,forecast the results of grey forecasting model by RBF Neural network,as the final forecast.Use the historical annual data of gas in one face for Fuxin,experimental results show that,in forecasting precision and training error,Gray-RBF model is better than a single gray model and the RBF neural network forecast model.This model count something simple,weakened the data randomness and model error,the forecasting average error of mine gas is 1.57%.