Identification of a research nuclear reactor using computational intelligence techniques
by Erick Rojas-Ramirez, Jorge Samuel Benitez-Read, J. Armando Segovia-De-los-Rios, Luis C. Longoria-Gandara
International Journal of Nuclear Knowledge Management (IJNKM), Vol. 5, No. 2, 2011

Abstract: In the field of nuclear engineering, the development of methods to identify the non-linear dynamic behaviour of nuclear reactors is an important area of research. In this paper, the design and performance of a multi-input multi-output Takagi–Sugeno (TS) neuro-fuzzy network are presented. The aim of this network is to identify the temporal behaviour of the power, the fuel temperature and the core reactivity of a TRIGA Mark III type nuclear research reactor. The tuning of the parameters corresponding to the antecedent membership functions is carried out by means of descent gradient algorithms with stable training, whereas the consequent parameters are identified using a Kalman estimator. Genetic algorithms are used to define the best input selection in the model. The results of the simulations show that the identification system converges rapidly and with high accuracy to both the training data and the test data. Thus, in the absence of real reactor data, the identified system can be used for tuning purposes of reactor power control schemes.

Online publication date: Wed, 18-Feb-2015

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Nuclear Knowledge Management (IJNKM):
Login with your Inderscience username and password:

    Username:        Password:         

Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email