论文部分内容阅读
提出利用LMD(Local Mean Decomposition)方法获取生产函数分量(PF分量)进行SVM(Support Vector Machine)建模,用此方法对采煤工作面瓦斯涌出量进行预测。通过LMD对瓦斯涌出量的历史数据进行分解得到其PF分量,然后,对应于每个PF分量各利用SVM函数拟合方法进行外推预测,再把不同PF分量的预测结果进行叠加重构合成,进而获得瓦斯涌出量预测的理论结果值。通过对某煤矿监测历史数据进行实例分析,可见此方法预测效果比常规SVM方法预测精度高,LMD的引入可大幅度提高瓦斯涌出量的预测精度,表明此方法建立的采煤工作面瓦斯涌出量预测模型具有较好的合理性和可靠性。PF分量的获取和SVM方法小样本预测的结合,能够充分发掘数据本身所蕴含的物理机制和物理规律,这也十分符合利用数据自身驱动来获取其影响因素相互间的物理机制,从而为瓦斯涌出量预测精度的提高奠定较好基础。
It is proposed that LMF (Local Mean Decomposition) method is used to obtain the production function component (PF component) for SVM (Support Vector Machine) modeling, and this method is used to predict gas emission from coalface. Through the LMD, the historical data of gas emission is decomposed to obtain the PF component. Then, corresponding to each PF component, the SVM function fitting method is used to extrapolate the prediction result, and the prediction results of different PF components are superimposed and recombined , And then get the theoretical value of gas emission prediction. Through the case analysis of a coal mine monitoring historical data, it can be seen that the prediction accuracy of this method is higher than that of the conventional SVM method. The introduction of LMD can greatly improve the prediction accuracy of gas emission. It shows that the coal seam gas The output forecasting model has good rationality and reliability. The combination of PF component acquisition and SVM small sample prediction can fully explore the physical mechanism and physical laws inherent in the data itself. This is also in line with the physical mechanism by which data are driven by each other to obtain their mutual influencing factors, Output forecasting accuracy to improve and lay a good foundation.