基于三维荧光和多传感器信息融合的水中嗅味物质快速检测方法研究

随着环境问题日渐突出，水环境保护和生态建设已经上升到国家战略发展高度。水体异嗅异味问题直接关系人民安全健康，引起了越来越广泛的关注。现有水中嗅味检测技术依赖人体感官、耗时长、成本高，难以满足水环境智能监测的要求。本项目以水中嗅味物质低成本、准确、快速检测为研究目标和重点，建立以嗅味物质三维荧光光谱特征提取为基础、结合多传感器信息融合技术的水中嗅味物质检测新理论、新方法。针对水环境复杂多变，提出基于多模型校正的光谱预处理及特征区域提取方法；运用先进信息处理技术，研究嗅味物质与光谱信息的关联关系，建立水中嗅味物质三维荧光特征模型及光谱指纹库；同时利用多传感器水质信息，研究嗅味物质与水质指标映射关系；综合利用三维荧光光谱及多传感器信息进行分布式、多层次融合，建立水中典型嗅味物质检测方法和系统原型。本项目作为环境、检测、信息等多学科交叉的前沿内容，为水环境监测和水污染防治提供理论、方法和技术支持。

With the increasingly prominent environmental problems, water environment protection and ecological construction have risen to the strategic height of the national development. The problem of taste and odor (T&O) pollution in water is directly related to people's safety and health, and has been arousing more and more widespread concern. Current T&O detection technologies either relies on human senses or complex instrument operation, which are affected by human factors, time-consuming and expensive. Consequently, they are difficult to meet the intelligent requirements in water environment monitoring. This project is aiming at developing a low cost, accurate and rapid detection technology of T&O substances in water. This innovative detection method is based on three-dimensional fluorescence spectral analysis and multi-sensor information fusion. Based on multi-model correction, the spectrum preprocessing and feature region extraction methods are studied and proposed, to reduce the impact on spectrum from complex and changeable water environment. The relationship between T&O substances and their spectral information are investigated by using advanced information processing technology. And three-dimensional fluorescence characteristic models and spectral fingerprint database of typical T&O substance in water will be established. Furthermore, water quality information obtained by multi-sensor is used to study the mapping relationship between T&O substances and multiple water quality indices. Multi-level information fusion methods are studied to make comprehensive use of three-dimensional fluorescence spectroscopy and multi-sensor information. Based on that, a low cost, accurate, rapid detection method of T&O substances in water will be established and the detection system prototype will be developed. As part of interdisciplinary in environment, detection technology and information, this project will provide theory, new method and technical support for water environment monitoring and water pollution prevention field.