Applications of Soft Computing in nuclear power plants: A review

• Published in: Progress in nuclear energy (New series) Vol. 149; p. 104253
• Main Authors: Ramezani, Iman, Moshkbar-Bakhshayesh, Khalil, Vosoughi, Naser, Ghofrani, Mohammad B.
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.07.2022
• Subjects: Conventional computational methods; Model independency; Nuclear power plants; Soft computing techniques; Soft computing techniques; Model independency; Conventional computational methods; Nuclear power plants; Data sets
• ISSN: 0149-1970
• DOI: 10.1016/j.pnucene.2022.104253

Soft Computing (SC) is defined as a group of computational techniques that solve complex problems independent of mathematical models. SC techniques including artificial neural networks (ANNs), fuzzy systems (FSs), evolutionary algorithms (EAs), etc., can solve problems that either cannot be solved by the analytical/conventional methods or require a long computation time. Due to their features, SC techniques are nowadays widely used in scientific and industrial researches. SC techniques have also been included in many types of research related to nuclear power plants (NPPs). In this paper, the applications of SC techniques in NPPs, according to published articles, are presented. The applications include fault detection and diagnosis, reactor power control, solving flow and heat transfer problems, optimization of NPP operation and design, High-performance accident management support tools, NPP operating parameters forecasting, and reliability/safety assessment of NPP. Other miscellaneous applications of SC that are used indirectly in NPP have also been mentioned. In addition, the advantages/disadvantages of SC techniques are compared with conventional computational methods. The main advantages of SC techniques are high execution speed, model independence, and high efficiency; while the main disadvantage is the need for access to reliable data sets which should be provided by experiment or modeling.