In the evolving landscape of energy production, nuclear power plants (NPPs) are facing increasing pressure to adapt and become more flexible to accommodate the growing share of renewable energies. Researchers Perceval Beja-Battais, Alain Grossetête, and Nicolas Vayatis from Framatome, a leading international nuclear energy company, have been working on enhancing the predictive capabilities of NPPs to meet this challenge.
In their recent research, the team focused on improving the Model Predictive Control (MPC) methods used in the Operator Assistance Predictive System (OAPS) developed by Framatome. MPC is a advanced control strategy that uses a dynamic model of a process to make predictions about future behavior, allowing for optimal control decisions. The researchers aimed to enhance these methods through data-driven simulation schemes.
The core of their work involved creating two surrogate models that act as alternative simulation schemes for nuclear reactor core simulation. These models are based on a set of nonlinear stiff ordinary differential equations (ODEs). The researchers demonstrated that both data-driven and physics-informed models can effectively integrate complex dynamics with a significantly reduced computational time—up to 1000 times faster than traditional methods.
This research was published in the journal “Annals of Nuclear Energy,” a reputable source for nuclear energy research. The practical applications for the energy sector are substantial. Faster and more accurate simulations can lead to improved operational flexibility, better integration with renewable energy sources, and enhanced safety and efficiency of nuclear power plants. By enabling more precise predictions and controls, these surrogate models can contribute to the overall stability and reliability of the energy grid, supporting the transition to a more sustainable energy future.
The work of Beja-Battais, Grossetête, and Vayatis highlights the potential of data-driven approaches in optimizing nuclear reactor operations, offering a promising avenue for the energy industry to meet the demands of a rapidly changing energy landscape.
This article is based on research available at arXiv.