A Review For Use of AI and ML Techniques On Nuclear Power Technologies For The Last Decade

Abstract

Research in the field of nuclear technology increasingly focus on artificial intelligence (AI) and machine learning (ML) techniques to make nuclear power plants easier to operate safer, and more reliable. This review investigates the integration of AI and ML in nuclear power technologies over the past decade. We collected 725 research articles from five leading journals related to nuclear technology, categorizing them into five distinct groups identified by the International Atomic Energy Agency (IAEA) based on their focus research areas. This study aims to investigate the evolution of research topics over the years. We also examined the keywords used within these studies to obtain insights into prevailing trends. Furthermore, we summarized the AI and ML techniques employed across these articles to understand their applications in the nuclear sector. This study demonstrates experience using artificial intelligence-supported methodologies to improve various aspects of nuclear technology and promote innovation in the nuclear industry. Over the last decade, the use of AI and ML in research on nuclear power reactors has significantly increased. In 2018, there was a rapid increase in research articles on AI and ML applications; this trend has increased linearly over the last five years. The groups with the largest share in the published articles are prediction and prognosis, analytics, and optimization, respectively. However, research articles about automation for nuclear power have been increasing significantly in the last five years.