Title :
Monitoring of continuous crystallization process
Author :
Aparna, A.M. ; Kundu, Madhusree
Author_Institution :
Dept. of Chem. Eng., N.I.T. Rourkela, Rourkela, India
Abstract :
Monitoring and control of continuous crystallization processes are highly challenging due to their nonlinear, multivariate and highly correlated nature. Present paper furnishes the successful implementation of various chemometric techniques like clustering of time series data and moving window based pattern matching aiming towards fault detection and classification of different operating conditions of a continuous crystallizer. PCA based combined similarity was chosen as the index of clustering as well as pattern matching.
Keywords :
chemical engineering; crystallisation; fault diagnosis; multivariable control systems; nonlinear control systems; pattern matching; principal component analysis; process control; process monitoring; time series; PCA based combined similarity; chemometric techniques; continuous crystallization process monitoring; continuous crystallization processes control; continuous crystallizer; fault classification; fault detection; moving window based pattern matching; time series data clustering; Crystallization; Crystallizers; Databases; Mathematical model; Monitoring; Pattern matching; Principal component analysis; PCA similarity; chemometrics; clustering; continuous; crystallizer;
Conference_Titel :
Control, Instrumentation, Communication and Computational Technologies (ICCICCT), 2014 International Conference on
Conference_Location :
Kanyakumari
Print_ISBN :
978-1-4799-4191-9
DOI :
10.1109/ICCICCT.2014.6993134
