盲信号分离技术在气体浓度分析中的应用研究-检测技术与自动化装置专业毕业论文.docxVIP

沈阳理工大学硕士学位论文 沈阳理工大学硕士学位论文 关键词：气体传感器；盲信号分离；混合气体分析；峭度；最大信噪比 沈阳理工大学硕士学位论文 Abstract As the advancement of the science and the development of industry, the requirements of the detection and analysis of multicomponent gases are on great rise. But existing detection methods cannot meet them. Therefore, combine the gas sensor array to Bland Signal Separation to analysis the consistency of mix gas such as high sensitivity, good selectivity, long-term stability, low cost and so on. Any type of gas sensor that can respond broadly to a range or class of gases rather than to a specific one can be employed in the detecting system. Because the infrared absorbed wave band of CO and CO2 have wraped, there is a error about the concentration. Analysing the detecting datas with the arithmetic of Blind Signal to advance the precision of detecting. The innovation of this paper is combining the arithmetic of Blind Signal Separation to the analysis of mixture gases. In this paper ,it introduces the basic model, principle of Blind Signal Separation, principal component analysis, independent component analysis, nonlinear principal component analysis, the arithmetic basised on the kurtosis and the arithmetic basised on the maximum Signal-to-Noise. Discussing the blinding separation and prep- disposaling the datas. Analysing the datas with the programme of these arithmetics in MATLAB. The resultes show that ICA and NLPCA are better than PCA in precision. ICA and NLPCA is more or less in the analysis of datas. But PCA is a means of pre-disposaling of signal. So in the basis of PCA, analyzing the detecting datas can increase the separated of mixture gases. Analysis six mixtures gases using the the arithmetic basised on the kurtosis and the arithmetic basised on the maximum Signal-to-Noise. The former does not estimate the pdf of two gases in advance. The average forecast error of CO and CO2 is 8.29% and 5.03%; The latter makes the best Signal-to-Noise as the activation function to calculate sepa