基于数据挖掘技术的瓦斯气体红外光谱定量分析方法的研究

瓦斯气体的监测是影响煤矿安全的重要因素之一，对其的在线实时监测成为煤矿安全的重要保障。本文采用红外光谱分析技术，对瓦斯气体的红外定量分析算法进行了研究。为提高模型分析精度，将数据挖掘技术应用在分析算法中。通过增加基于K平均非分层聚类分析对数据进行处理发现，带聚类分析的偏最小二乘算法比单纯采用偏最小二乘算法在精度上明显占优。另外，为减少模型中个别定标样本误差对精度的影响，采用聚类分析的方式进行了数据预处理，发现这种去噪方法在分析精度上又有所提高。

Study of Infrared Spectroscopy Quantitative Analysis Method for Methane Gas Based on Data Mining

Monitoring of methane gas is one of the important factors affecting the coal mine safety. The online real-time monitoring of the methane gas is used for the mine safety protection. To improve the accuracy of model analysis, in the present paper, the author uses the technology of infrared spectroscopy to study the gas infrared quantitative analysis algorithm. By data mining technology application in multi-component infrared spectroscopy quantitative analysis algorithm, it was found that cluster analysis partial least squares algorithm is obviously superior to simply using partial least squares algorithm in terms of accuracy. In addition, to reduce the influence of the error on the accuracy of model individual calibration samples, the clustering analysis was used for the data preprocessing, and such denoising method was found to improve the analysis accuracy.