复杂对象高维模式识别方法及其食品精准鉴别创新性研究

In the studies on precise verification of food, the difficulty and major breakthrough is to develop a rapid, effective and smart method based on high-way pattern recognition. In this project, for the imperfection of traditional targeted analysis and non-targeted fingerprinting in the food authentication, we try to take advantage of “mathematical separation” and “second-order and higher-order advantages” in the higher-dimensional algorithms developed by our group, with the objectives to develop multiple linear decomposition algorithms and high-way pattern recognition methods based on liquid chromatography coupled with mass spectrometry (LC-MS), excitation-emission matrix fluorescence and so on. The methods developed by our group also have advantages in the fast, direct, accurate and precise quantification for targeted analysis and in the “mathematical separation” and resolution for non-targeted analysis. By virtue of this strategy, we can further solve problems such as screening and discriminating special biomarkers in complex food system and identification of genetically modified food. Based on the theory and application of multi-way resolution and calibration developed by our group, we will explore new methods which will solve the problems in multiple targeted analysis and non-targeted fingerprinting in the food authentication. And the researches will provide new tools for the precise verification of food and further contribute to the development of high-way pattern recognition methods. What is more, food quality and safety will be further guaranteed, and the consumer rights and interests will also be strengthened.

发展快速、高效、灵巧的高维模式识别方法是解决当前食品等物类真伪精准鉴别研究的难点、突破口和重要组成部分。本项目拟针对传统目标食品分析方法和非目标食品指纹识别方法在食品等真伪鉴别中不足，利用本室前期发展的高维算法独具的“数学分离”和“二阶及高阶优势”特性，及其在目标分析体系中的精准定量和非目标分析体系中的数学分离及分辨的综合优势，发展建立基于液质联用和三维荧光等量测手段的多线性分解和高维模式识别新方法，并进一步解决复杂食品体系等的特征生物标记物的筛选和识别及转基因食品的鉴别等食品真伪鉴别难题。本研究将基于本室近二十年来对化学多维分辨及校正基础理论及应用研究，瞄准食品质量安全和食品真伪难辨等物类判别难点，开展化学计量学中高维模式识别方法前沿研究，期望实现基于多目标食品分析和非目标食品指纹识别的高维模式识别方法创新，为食品等物类真伪精准鉴别提供新工具，为进一步发展高维模式识别方法做出新贡献。

发展快速、高效、灵巧的高维模式识别方法是解决当前食品等物类真伪精准鉴别研究的难点、突破口和重要组成部分。本项目针对传统目标食品分析方法和非目标食品指纹识别方法在食品等真伪鉴别中不足，利用本室前期发展的高维算法独具的“数学分离”和“二阶及高阶优势”特性，及其在目标分析体系中的精准定量和非目标分析体系中的数学分离及分辨的综合优势，发展建立基于液质联用和三维荧光等量测手段的多线性分解和高维模式识别新方法，并进一步解决复杂食品体系等的特征生物标记物的筛选和识别及转基因食品的鉴别等食品真伪鉴别难题。本研究将基于本室近二十年来对化学多维分辨及校正基础理论及应用研究，瞄准食品质量安全和食品真伪难辨等物类判别难点，开展化学计量学中高维模式识别方法前沿研究，期望实现基于多目标食品分析和非目标食品指纹识别的高维模式识别方法创新，为食品等物类真伪精准鉴别提供新工具，为进一步发展高维模式识别方法做出新贡献。本项目发表多篇中英文高水平综述文章，发表SCI源刊物学术论文近30篇，项目执行期间，参加国内外学术会议数次，发表论文近十篇次，同时培养毕业硕士研究生8名，同时培养毕业博士研究生4名。取得很好的社会效益及科学进步。