小波降噪技术在差分吸收光谱浓度检测中的应用

为了实现对工业气体SO₂的浓度进行监控，基于紫外差分吸收光谱法开发了SO₂在线检测系统。针对系统噪声和Mie散射使吸收光谱叠加带来的误差，本文提出采用小波变换降噪技术代替传统光谱处理方法中的多项式平滑滤波技术来提高检测精度。通过对应用了Symlets、Daubechies、Coiflet和Biorthogonal这4种不同小波函数的实验数据分析和对传统小波阈值选取方式的改进，最终确定了基于rigisure阈值的小波阈值去噪的信号处理方法，并提出一种新的信噪比量来衡量信号处理的效果。这种方法可以快速可靠地处理光谱信号，处理后所得的监测浓度准确度基本控制在1.5%以内。在实验室环境下和工业现场环境下的大量实验结果表明本方法能有效的减小噪声对SO₂浓度监测带来的影响。

Application of Wavelet De-noising Method in Concentration Measuring System Based on DOAS

The concentration measuring system based on DOAS is developed to watch the concentration of the sulfur dioxide in pollution gases. In order to decrease the influence of the system noise and the Mie Scattering,the wavelet de-noising method is used. By comparing 4 kinds of different wavelet function and improving traditional threshold value, the de-noising method based on rigisure threshold is found, and a new measurement is put forward to measure the effects of signal processing. This kind of technology is applied to processing the spectral signal accurately and reliably, and the error is controlled under 1.5%. Experimental research results and industry scene results show that the wavelet de-noising method can greatly improve the SNR of the system and minimize bad effects of noises.