Identification of a Continuous Stirring Tank Reactor Using Two-layer Perceptron Neural Network Model

Abstract

This paper investigates the identification of a continuous stirring tank reactor using a two-layer multi-layer perceptron neural network. It is shown that the neural network based state-space innovations form model clearly outperforms the equivalent linear model in simulation performance and correlation functions of prediction errors and thus, it is possible to obtain hopeful results for this system with neural network based state-space innovations form model. It is by no means claimed that the “optimal” solution is found, however, it is shown that the proposed neural network based model has not only significantly performed better in simulation but also in correlation characteristics than the equivalent linear model in the sense that the correlation functions tend to lie more closely within the confidence regions. This also means that most of the information has been extracted from the training set and that the model approximates the given system well.

Indexing Terms:

• Levenberg-Marquardt Algorithm
• Multi-layer perceptron neural network
• Neural network based state-space innovations form model
• State-space innovations form model
• System identification

Additional information

Notes on contributors

S V Dudul

S V Dudul was borne on 28th of August 1964. He received his B.E. (Electronics & Power Engg.) from Government College of Engineering, Amravati in 1986 with distinction and gold medal, ME (Electronics with Specialization in Computer Applications) from Walchand College of Engineering, Sangli in 1989 with distinction and PhD in Electronics Engineering from Amravati University, Amravati in 2003. Presently, he is working as Head of the Department of PG Department of Applied Electronics, Amravati University. His fields of interest are Engineering applications of Soft-computing techniques, Machine learning, System Identification and Signal Processing. He has over 38 publications to his credit in National and International Journals and Proceedings of the International conferences. He has been invited to work as a reviewer for several papers submitted to the Elsevier International Journal of Applied Software Computing. He has conducted numerous comprehensive workshops on Nero Computing techniques and their applications at various engineering colleges. He also chaired International as well as National conferences. Presently, he is guiding numerous students for their doctoral degree. He is also instrumental in the execution of some UGC projects. He is a Fellow of IE (India) and IETE. He is also a chartered engineer of IE (India). He is a member of IE (India), ISTE, IETE, CSI, ISA, and ISCA.

A A Ghatol

A A Ghatol was born on 29th of August 1948, obtained his MTech (Elect Engg) from IIT, Mumbai in 1973 and PhD from the same institute in 1984. He was a Principal of Government College of Engineering, Amravati and later Pune Institute of Engineering and Technology, Pune (formerly known as Govt College of Engineering, Pune) from 1991 to 2005. In April 2005, he became Vice-Chancellor of Dr Babasaheb Ambedkar Technological University, Lonere (Dist. Raigarh). He has guided over thirty research students for their PhD degree and has more than 50 publications in national and international journals to his credit. He has successfully completed many research projects sponsored by AICTE, DST World Bank and CSIR. He was instrumental in getting autonomy to Govt College of Engineering, Pune. He is a fellow of IE (India), ISTE, ISA, and a member of IEEE.