面向水质多参数在线检测的紫外-可见光谱影响机理及提升测量稳定性方法研究

In order to solve the key problems in the on-line detection of multi-parameters of water quality by ultraviolet-visible spectroscopy, such as interference of instrument and external stray light and noise, difficult to model the absorption spectrum analysis due to complicated water quality components, parameter measurement of spectral interference caused by scattering of suspended particles in water, the project research on the measurement stability and analytical stability of water spectrum detection, the mechanism of scattering on water absorption spectra and water quality parameters. Aiming at solving the problems of instrument noise, stray light noise and photoelectric conversion efficiency existing in the optical measuring unit of water quality, the work of research on improving the measurement stability is carried out. In particular, differential optical path is applied to the design of optical path of water quality testing, the measurement mechanism is explored, and the differential data processing method is explored. The mechanism and mechanism of light scattering by suspended particles in water are explored to provide theoretical basis for the data processing of water quality spectrum. The pretreatment method based on water quality spectral data is studied to improve the analysis precision and the stability of measurement. The universal water quality multi-parameter measurement model is explored, and the water sample modeling adaptive method for different test environments and different water bodies is studied. The development of the project can help to improve the multi-parameter measurement and analysis stability of water quality. The research mechanism, model, theory and method can also be widely applied to the environmental spectrum analysis in the atmosphere or soil, which has important theoretical significance and application value.

项目针对紫外-可见光谱法水质多参数在线检测中存在仪器及外界杂散光噪声干扰大、水质成分复杂导致吸收谱难以建模分析、水中悬浮颗粒物散射引起光谱干扰影响参数测量等关键问题，开展水质光谱检测的提升测量稳定性及分析稳定性方法研究，及散射对水质吸收光谱及水质多参数的影响机理研究。针对水质光学测量单元存在仪器噪声、杂散光噪声严重等问题，开展提升测量稳定性的研究工作，特别地，提出差分光程应用于水质检测光路设计，研究其测量机理，探索差分数据处理方法；探索水中悬浮颗粒物对光散射的影响机理，为水质光谱的数据处理提供理论依据；研究基于水质光谱数据的预处理方法，以提高分析稳定性；探索建立普适性水质多参数测量模型，研究针对不同测试环境、不同水体的水样建模自适应方法。项目的开展有助于提升水质多参数测量及分析稳定性，研究的机理、模型、理论及方法亦可推广应用于大气或土壤等环境光谱分析，有重要的理论指导意义和的应用价值。

紫外-可见光谱法检测水质多参数具有无需化学试剂、无二次污染、分析速度快、实时在线、体积小、成本低、可实现多参数测量等优点，针对紫外-可见光谱法水质检测的仪器、散射影响、数据预处理与多参数解算中存在的亟待解决的科学问题，本项目就：提升测量稳定性的水质吸收光谱检测方法研究，水中悬浮颗粒物散射对水质紫外-可见吸收光谱的影响机理研究，提升分析稳定性的水质光谱数据预处理及多参数解算方法研究，三个方面进行深入研究和探讨。完成了：基于紫外-可见光谱分析技术的水质多参数在线检测系统搭建，设计并加工制作了水质光谱检测单元，数据传输单元、水污染云监测平台；研究了水中不同种类、形状悬浮颗粒物与内部光传递的作用关系，建立水中悬浮颗粒物的散射模型，提出了运用压缩感知理论，对水质浊度进行校正的方法；提出基于EWMA-PCA的归一化算法在水质光谱数据标准化中的应用研究，有效地解决了不同仪器间的数据标准化和模型传递问题；提出了将遗传算法应用于小波阈值优化的去噪方法，基于压缩感知的去噪方法，提高了信噪比，有效地抑制了噪声干扰。提出了基于压缩感知的数据校正方法，对水质光谱信号进行了有效的校正处理；提出采用深度学习方法，对水质进行分类和预警研究。项目形成了科研论文和专利等成果：发表论文13 篇，其中包括SCI 检索论文2 篇，EI 检索论文2篇，核心期刊3 篇，会议论文4 篇。申请发明专利4项（已公开），获授权实用新型专利1 项。培养研究生9名，参加国际国内学术会议4 次，完成了项目任务书规定的目标。项目的实施对光谱法水质多参数检测有着重要的理论指导意义，同时，借鉴、推广江浙经验，为“河长制”推行贡献力量。相关研究成果亦可推广应用于大气污染和土壤污染检测领域，形成的环境光谱分析理论、仪器、方法及技术成果，推动环境在线检测的产业化升级，必将改变行业现状，引领环境监测领域的技术革命。