A Bayesian Neural Network (BNN) Machine Learning (ML) Surrogate Modelling Framework for High-Fidelity Thermal Fatigue Modelling of Components

点击翻译按钮获取中文摘要

National energy grid and power distribution infrastructure, within developed and developing countries, are currently undergoing significant modification to produce so-called smart grid infrastructure. In these smart grids there will be a mixture of dynamically evolving or intermittent renewable energy supply augmented by a proportion of base-load electrical power. There is a widespread misconception that nuclear power plants (NPPs) can only provide relatively inflexible base-load power to any national energy grid networks. However, this is significantly affecting the long-term prospects of the nuclear sector's role in delivering cost effective electrical power within the more heterogeneous and mixed power generation environment that is evolving around the world. Increasing electrical power generation is being dominated by intermittent renewable energy forms. Nuclear power generation will need to adapt to this new smart grid infrastructure.