نوع مقاله : مقاله پژوهشی
نویسندگان
دانشکدهی مهندسی انرژی، دانشگاه صنعتی شریف، صندوق پستی: 1114-14565، تهران - ایران
کلیدواژهها
عنوان مقاله English
نویسندگان English
This paper presents the design of an artificial neural network (ANN) to identify the In-Core Loose Parts in the Instrumentation Position (ILOPAIP) in the Bushehr nuclear reactor. Initially, a neutron noise simulator generates the outputs for the 54 neutron detectors in the reactor core due to ILOPAIP. These outputs are then used to design and train a multilayer perceptron. The neutron noise source caused by ILOPAIP comprises two components: an absorber of variable strength type. Therefore, the first step involves determining the location of the neutron noise source of the absorber type with variable strength. An ANN with a hidden layer is designed to identify the location of this neutron noise source. The results show a deviation of about 1% from the desired output values. Subsequently, the location of ILOPAIP is identified using an ANN with two hidden layers. The results show a deviation of about 3% from the desired output values. Overall, the findings indicate that due to the complex relationship and the dependence of neutron noise on the location of the detectors, the use of ANN is effective for this identification.
کلیدواژهها English
10. Yegnanarayana B. Artificial Neural Networks. New Delhi: Rajkamal Electric Press. 2005.