The Application of Artificial Intelligence to Nuclear Power Plant Safety

Through application of artificial intelligence (AI), the burden of analytical computational load for analysis of any given problem where countless variables have to be taken into account, is virtually eliminated. Since for engagements in real life operations and instantaneous actions are of paramount importance and vital, AI can be a strong alternative to overcome the complex problem solving in short time frames. As such, in this study a brief review of AI basics is given and literature for AI applications in nuclear field such as defect detection in nuclear fuel assembly, dose prediction in nuclear emergencies, fuel and component failure detection, core monitoring for reactor transients, core fuel optimization models, gamma spectroscopy analysis and specifically nuclear reactor safety in operation are assessed. Afterwards, an AI model for analyzing transients in VVER type nuclear power plants that is being built in Turkey is proposed. This model must keep up with instantaneous data flow and giving actionable feedback to the operator both for the cause and the solution. A semi-autonomous AI control system that help the operator decision making is a significant contributor to the safety of a reactor.