Web-based Chromatography System Development And Chromatographic Analysis Algorithms Research

Chromatographic analysis is a technique for separating the components of a mixture.It is currently widely used in many fields such as chemical engineering,biology,and pharmaceuticals.Chromatographs are devices used for chromatographic separation and analysis,it include many devices such as detectors,chromatographic pumps,and temperature/pressure controller.Recent years,the development of computer technology has promoted the automation and intelligence of chromatographic analysis.The steps of chromatographic analysis are gradually developed to be controlled by chromatographic analysis software,which form as an intelligent chromatography system.This paper uses web technology to design and implement a chromatography analysis system.Users can control various chromatographic equipment through this system and monitor their status in real time.At the same time,they can perform data processing on the collected spectrum peak signals to obtain the results of chromatographic analysis.This system adopts the architecture of browser + server + chromatography equipment.The browser is an interface for human-computer interaction,which mainly implements the rendering of the virtual instrument monitoring panel and chromatographic signals.The server is connected to the chromatography equipment.According to the different requirements of the system functions,the server is divided into web services,instrument communication services and chromatography algorithm SDK.Web services are used to respond to user requests and provide basic application functions,while instrument communication services are responsible for communicating with chromatographic instruments.The separation between the two services improves system stability,and a three-level message format is used for data exchange.A chromatography algorithm SDK is implemented by concurrency model to provide system data analysis functions,which support multi-task operation at the same time.Chromatographic signal processing is the core module of the system,including steps such as signal denoising,chromatographic peak identification,overlapping peak decomposition,and area calculation.This article first introduces commonly used methods of chromatographic signal denoising,then implements a filtering method combining the mean and wavelet-soft-threshold based on the characteristics of actual chromatographic data,which can improve signal-to-noise ratio while maintaining peak shape.In addition,the finite state machine(FSM)model of the first derivative chromatographic peak detection method is established,which can effectively identify the characteristic points of the chromatographic peak automatically.Finally,aiming at the problem of overlapping peaks in chromatographic analysis,this paper studies and implements a method of fitting sub-peak area ratios with inflection point information using a random forest model,the model training is efficient while the sub-peak decomposition fit is accurate.Then a chromatographic overlapping peak analysis algorithm based on particle swarm search for optimal parameters is implement,which can decompose chromatographic multi-peak overlap.Besides,the results of these two methods are discussed and compared.Finally,the function and performance of the whole system are verified and tested in this paper.The system has good performance,and the algorithm has a fast calculation speed,supports multi-task running at the same time,which can identify the chromatographic peaks in the signal and effectively resolve the overlapping peaks.The results show that the system meets the requirements of actual use.