红外光谱分析技术是应用前景非常广阔的过程分析技术之一。利用原位红外技术对反应过程进行跟踪，可以精确地监测反应的起点、终点、不稳定中间体的生成与消失等情况，采集到的二维光谱矩阵中包含了大量的反应动力学信息。高精度的反应动力学模型是制药反应器设计、制药工艺放大以及制药生产过程控制的重要基础。基于传统的在线采样和离线分析技术进行反应动力学参数估计已经很难满足现代制药生产的要求。本项目将聚焦于基于在线红外光谱的反应动力学参数估计过程中的理论和技术难题，通过结构增广矩阵的零空间分析、干扰光谱的双线性分解、灵敏度分析、赤池信息准则、极大似然估计、回溯同伦等原理、方法和技术对多因素影响下的反应动力学参数估计问题、反应动力学参数的可估性和估计精度问题、联立框架下反应动力学参数估计求解问题等关键问题进行深入研究，为反应动力学模型在实际中的应用打下坚实的基础。

Infrared spectral analysis is one of the most promising process analysis techniques. The reaction process can be tracked by the in situ infrared technique. The beginning and the end of the reaction, as well as the generation and the disappearance of the intermediate can be exactly monitored. The collected two-dimensional spectra matrix contains abundant kinetic information of the reaction. High precision reaction kinetics model plays an important role in the reaction design, scale-up calculations from laboratory to commercial, and process control for pharmaceutical production. The traditional online sampling and offline analysis technique based kinetic parameter estimation cannot satisfy the requirement of modern pharmaceutical production. In this proposal, we plan to focus on the theory and technology difficulties appeared in the online infrared spectroscopy based reaction kinetic parameters estimation. The key fundamental problems including kinetic parameters estimation influenced with multiple unwanted factors, estimability analysis of the kinetic parameters and precision analysis of the estimated parameters, and solving problem for kinetic parameters estimation based on simultaneous approach, will be further researched based on the principles, approaches and technologies consisting of null space analysis for augment matrix of reaction model structure, bilinear decomposition for spectra disturbance, sensitivity analysis, Akaike information criterion, maximum likelihood estimate, backtracking homotopy and so on. It is hoped that a firm application foundation for the reaction kinetics model can be established by this research.