封装西林药瓶残留氧气检测中的谐波基线校正和去噪方法

摘要

For the proven technology and fast noncontact advantages of gas concentration measurement,the tunable diode laser absorption spectroscopy(TDLAS)is very suitable for the residual oxygen concentration detection in the packaged Xilin bottle.In this paper,TDLAS is used for residual oxygen detection in the packaged xilin bottle.The detection light goes through the air and glass bottle,and the scattering and attenuation to the glass wall of the laser are the main interferences which have a great impact on the stability of the second harmonic signal.This paper designed and built the residual oxygen detection system in packaged bottle based on TDLAS.According to the second harmonic signal collected from the system,a new method based on wavelet transform is proposed and the effect is obvious.Firstly,using "sym6" wavelet,the measured signal was decomposed by five layers of wavelet,and the corresponding baseline slope was obtained according to the low frequency component of each layer.Then,the baseline slope of the original signal was obtained by weighted average of the five baseline slopes.According to the obtained baseline slope,the original signal was processed by the baseline,and the reconstructed signal was obtained after the wavelet decomposition and the soft threshold processing.The measurement results of the bottle with 21％oxygen concentration showed that the relative error between the treated and the theoretical signals decreased from 1.26％to 0.12％,which proved that this method could solve the problems of baseline drift and noise interference in the process of residual oxygen concentration detection,and overcome the interference of the glass wall to the second harmonic signal,providing a high quality signal for measuring oxygen concentration.%可调谐激光二极管吸收光谱(TDLAS)技术的成熟和快速非接触气体浓度测量的优点,十分适合用于对封装西林药瓶内残留氧气进行浓度检测.采用TDLAS技术对封装西林药瓶内残留氧气进行浓度检测,检测系统的光路经过空气和玻璃药瓶,玻璃瓶壁对激光的散射和衰减是检测系统的主要干扰,给二次谐波信号的稳定性带来了很大影响.设计和搭建了基于TDLAS的封装西林药瓶残氧量检测系统.针对从系统中提取出的二次谐波信号,提出了一种基于小波变换的基线消除和噪声滤除方法,解决在残留氧气浓度检测过程中基线漂移和噪声干扰问题,克服玻璃瓶壁对二次谐波信号的干扰,效果明显.选用"sym6"小波,将实验测得的信号进行五层小波分解,根据每一层小波分解得到的低频分量求出相应的基线斜率,对五个基线斜率进行加权平均得到原始信号的基线斜率.由求得的基线斜率,对原始信号经过去基线处理,再进行小波分解和软阈值处理后得到重构信号.对氧气浓度为21％的西林瓶的测量结果表明,处理后谐波信号和理论信号之间的相对误差由处理前的1.26％下降到了0.12％,证明了此方法可以很好地解决在残留氧气浓度检测过程中基线漂移和噪声干扰问题,克服玻璃瓶壁对二次谐波信号的干扰,为氧气浓度测量提供很高质量的信号.