Group method of data handling and neural networks applied in temperature sensors monitoring

Abstract

In this work a monitoring system is developed based on the Group Method of Data Handling (GMDH) and Artificial Neural Networks (ANNs) methodologies. GMDH creates non-linear algebraic models for system characterisation and ANN is a massively parallel distributed processor made up of simple processing units called neurons. The monitoring system was applied to the IEA-R1 research reactor at Instituto de Pesquisas Energeticas e Nucleares (IPEN) by using a database obtained from a theoretical model of the reactor. The IEA-R1 research reactor is a pool-type reactor of 5 MW cooled and moderated by light water, and uses graphite and beryllium as reflector. The two methodologies (GMDH and ANN) were combined to develop a temperature monitoring system. The results were compared with previous works where GMDH and ANN were used separately and the results obtained showed an improved monitoring system.