基于小波分析的GM(1,1)模型在瓦斯浓度预测分析中的应用

Application of GM( 1,1) model based on wavelet analysis in gas concentration prediction analysis

针对含噪声的煤矿瓦斯浓度时间序列,传统静态灰色GM(1,1)预测模型容易受到随机干扰影响致使预测精度不高的问题,建立了基于小波消噪技术的动态等维灰色理论预测模型,并以新信息优先设置模型初始值,克服了传统静态灰色理论模型在预测过程中不考虑未来因素对系统影响的缺陷。实例分析表明,文章建立的消噪动态灰色模型预测效果远优于传统静态灰色模型,预测结果更优,精度更高。

According to the time sequence of coal mine gas concentration with noise,the traditional static grey GM（ 1,1） prediction model is easily affected by the random disturbance resulting prediction accuracy is not high. To establish dynamic equal dimensional the grey theory prediction model based on wavelet denoising technique,and with new information priority setting initial value of the model,to overcome the traditional static gray theory model does not consider the future impact on the system defect factors in the prediction process. Example analysis shows that the established dynamic grey model to forecast the de- noising effect is much better than the traditional static gray model,prediction results better,higher accuracy.