Artificial intelligence and machine learning applications in the Spanish nuclear field

Abstract

Machine Learning and Artificial Intelligence techniques are increasingly applied in the nuclear field. In this way, Spanish companies and research institutions have been developing different projects in the field of nuclear energy applying such techniques. A brief overview of some of these projects is presented, which cover areas as diverse as nuclear data, nuclear fuel manufacturing, simulation codes, fuel performance monitoring, nuclear inspection, and operation training. Among the techniques used to develop them are deep neural networks for building surrogate models, convolutional neural networks for computer vision, Bayesian networks and Gaussian processes for predictive models, and speech recognition with large language models.