混合物质远程时域拉曼二维相关谱多维解析方法研究

Time-domain Raman spectroscopy technology has the ability of stand-off in-situ detection, and has broad application prospects in deep space surface material detection, polar-ice exploration, safe distance detection of dangerous drugs, etc. However, at present, Raman spectroscopy technology lacks effective spectral information extraction and identification methods, resulting in its inability to identify the mixture components in complex environments. In order to improve the recognition ability of time-domain Raman spectroscopy and break through the technical barriers in low accuracy, slow speed and anti-noise of the existing spectrum recognition methods for complex mixtures, this project combines the advantages of time-domain Raman photons to propose a multi-dimensional analytical method on two-dimensional correlation spectroscopy of mixtures for effectively extract spectral information. The main research contents and innovations include: based on the parameter down conversion and path integration theory of light-matter interaction, the photon fluctuation and correlation characteristics of Raman signal in time domain are studied; the mathematical model of Raman spectral information de-noising extraction and multi-dimensional decomposition based on photon correlation is established to realize the identification and quantitative characterization of diverse and minimal material components; and the establishment of remote Raman measurement in time domain. Based on the built timedomain Raman experimental system, algorithm optimization and performance evaluation are carried out . Through the research of this project, the practical level of long-range time-domain Raman technology in complex environment such as deep space exploration can be greatly improved, and it has positive scientific significance for the development of new spectral recognition technology.

时域拉曼光谱技术具有远程原位探测能力，在深空表面物质探测、极地冰矿勘探、危险药品安全距离检测等方面具有广阔应用前景，但是目前拉曼光谱分析技术缺乏有效光谱信息提取和识别方法，导致其无法识别复杂环境下混合物质成分。为提高时域拉曼光谱识别能力，突破现有谱图识别方法对于复杂混合物质成分识别准确率低、速度慢、抗噪差的技术壁垒，本项目结合时域拉曼光子特性，提出混合物二维相关光谱多维解析方法，有效提取光谱信息。主要研究内容和创新包括：基于光与物质相互作用的向下参量转换和路径积分理论，研究时域拉曼信号光子涨落和关联特性；建立基于光子关联的拉曼光谱信息降噪提取和多维分解的数学模型，实现多样化，微量化的物质成分识别和定量表征；创建时域远程拉曼测试系统，开展算法优化和性能综合评估。通过本项目研究，可以提升远程时域拉曼技术在深空探测等复杂环境中的实用化水平，对于发展新型光谱识别技术具有积极科学意义。

目前拉曼光谱分析技术由于信号弱和受荧光和背景光干扰等弱点，缺乏原位工作能力，极大影响该技术的应用场景。为提高时域拉曼光谱识别能力，突破现有谱图识别率低、速度慢、抗噪差等技术壁垒。本项目提出新型时域远程拉曼光谱技术，通过搭建并优化时域远程拉曼光谱探测实验平台，开展了时域拉曼系统光子涨落和关联特性的研究，提出基于噪声和信号关联特性的多次迭代提高信噪比的方法和二维相关谱矩阵的拉曼光谱信息提取模型，以及建立基于主成分分析和神经智能网络识别的多维解析方法，将复杂物质成分的拉曼光谱进行探测分析，同时开展时域拉曼实验验证及性能综合评估，对不同化学物质，低维二维材料晶体构型和医学活体组织进行有效识别检测。本项目研究成果有助于提高拉曼原位探测能力，在生化检测，边境安防，材料物理等领域有重要应用前景。

内容获取失败，请点击重试