中文摘要：

结合盲分离算法和热催化传感器特性，对混合气体进行分离，讨论了混合气体分析的盲可分离性。使用一个热催化传感器，引用自然梯度的盲分离算法对CH4和CO的混合气体进行分析和验证。实验结果表明，此方法可有效地分析CH4和CO的浓度，提高了分析精度。

英文摘要：

Combing blind source separation algorithm and the characteristics of heat catalytic sensor, the gas mixtures is separated. The blind signal separability of the gas mixtures is discussed. The gas mixtures of CH4 and N2A are separated by using a heat catalytic sensor and the results are analyzed by the algorithm based on the Natural gradient blind signal separation. The results show that this method is effective in analyzing the concentration of CH4 and CO and enhance the analysis precision.