Fault Isolation Using a Reconstruction Algorithm : By Using Auto-Associative Neural Networks

ISBN-13: 9783659323843 / Angielski / Miękka / 2013 / 88 str.

Sayyed Hamidreza Mousavi; Mehdi Shahbazian

Opis

Process history based approaches for fault diagnosis has been widely used recently. Principal Component Analysis (PCA) is one of these approaches, which is a linear approach; however most of the processes are nonlinear. Hence nonlinear extensions of the PCA have been developed. Nonlinear Principal Component Analysis (NLPCA) based on the neural networks is a common method which is used for process monitoring and fault diagnosis. NLPCA based neural networks are implemented using different methods, in this book we apply Auto-Associative Neural Networks (AANN) for implementing NLPCA. This work is aimed towards the development of an algorithm used in conjunction with an Auto Associative Neural Network (AANN) to help locate and reconstruct faulty sensor inputs in control systems. Also an algorithm is developed for locating the source of the process fault.

O autorze

Sayyed Hamidreza Mousavi Received his master s degree in Instrumentation & Automation Engineering and his B.S. degree in Petroleum Engineering (Technical Inspection) from Abadan Institute of Technology(AIT). He is currently a research fellow at Petroleum University of Technology focusing on Process monitoring and Nondestructive testing