一种用于海水DOC微弱信号去噪处理的小波多阈值算法研究

Research on Wavelet Multi Threshold Algorithm for Denoising Weak Signal of DOC in Seawater

针对目前国内尚无海水溶解性有机碳(DOC)现场自动化分析仪器的现状,本文基于臭氧氧化有机物发光的原理,针对光电检测过程中噪声信号的影响,提出一种基于小波多尺度多阈值的微弱信号去噪处理算法。该算法采用一种改进型的阈值函数,通过选择合适的小波基,灵活设置软硬阈值进行信号去噪。经过分析测试系统长期海试数据表明,该方法测量数据与实验室测量数据的相关性可达78.5%,证明了该方法稳定可靠、具备在线分析的能力,具有良好的推广价值。

In view of the current situation of the lack of on-site automatic analysis instruments for seawater dissolved organic carbon(DOC)in China.Based on the principle of ozone oxidizing of organic matter luminescence,aims at the influence of noise signal in photoelectric detection process,a weak signal de-noising algorithm based on wavelet multi-scale and multi threshold is proposed.The algorithm adopts an improved threshold function,and flexibly sets soft and hard thresholds for signal de-noising by selecting a suitable wavelet base.The long-term sea trial data of the test system show that the parallelism of the measurement data can reach 78.5%of the correlation with the laboratory measurement data,which proves that the method is stable and reliable,has the ability of on-line analysis,and has good popularization value.