Development Of Algorithms Library Of Industry Online-chromatographic Work Station

Traditional chromatographic analysis usually lags behind the process of production, can not be used in regulating and monitoring of production in time. The development of industrial automation requires the realization and popularization of online chromatography in order to achieve real-time analysis and control. The rich library of chromatography algorithms can provide optimal algorithm for the system and enhance the versatility of online chromatography system. On the understanding of the hardware and software structure of online chromatography system, the thesis develops online chromatography algorithms and builds a library of algorithms framework to achieve the scalability of the algorithm library.On the background of understanding the online chromatography system, the thesis studies the online chromatography system algorithm and develop algorithms’library. After brief introduction the need of filtering noise and kinds of noise, the thesis implements kinds of filtering methods and takes a online chromatography system which detects of kinetic parameters of PVC polymerization system for example to analysis and compare the methods of filtering noise, and provides the basis for choices of different situations. On the basis of analysis of the characteristics of online chromatography system,the author develop a variety of online peak identification methods, including the first derivative of deformation method, ruler method and curve fitting method, these methods take advantage of the stability of the detected substances, overcome the instability of baseline caused by unstable environment, and improve the automation of the online chromatography algorithm. The thesis discusses the process of computer implementation of the first derivative of deformation method in-depth and uses it to detect peak of liquor sample. In the end, through the discussion of the similarities and differences of the peak identification methods, the author extract the core of the peak identification process, and then build a algorithms’ library which be scalable and can chooses the best adaptive algorithm for systems. And also the thesis shows the interface of the algorithm library.And at last test the algorithm library.