基于中位值和小波两级滤波的气相色谱分析仪数据预处理方法

Median and wavelet two-stage filters based pretreatment method for gas chromatography analyzer data

当采用气相色谱法对混合气体进行分析时,由于存在奇异噪声和高斯噪声,对色谱峰检测产生不良影响。为此,通过小波变换的方法提取噪声,对噪声进行分析,建立噪声的概率分布模型,确定色谱信号中噪声的特点。根据噪声的特点,采用中位值滤波和小波变换相结合的滤波方法,处理色谱信号。实验表明,中位值滤波和小波变换相结合的两级滤波方法能有效滤除奇异噪声和高斯噪声,且信号畸变小,为后续色谱峰检测、重叠峰分离打下良好的基础。

It is singular noise and Gauss noise that have impact on peak detection when the gas chromatograph method is used to analyze the products contents. Aiming at solve this problem,the noise distribution model is established after analysis of the noise,which is extracted by wavelet transform. Considering the features of noise,the two-stage filter which combines median with wavelet is employed. The filtering method is effective,and the signal distortion is small. It benefits peak detection and overlapping peaks separation.