Sensor Optimization of Nuclear Reactor Subsystems within a Digital Twin Network

Nuclear power plants require continuous monitoring of systems, structures, and components for safe and efficient operations. Critical safety testing of new fuel compositions and power transient analysis on core temperatures are achieved through modeling and simulations, capturing dynamics associated with failure modes to create digital twins. Accurate reconstruction of temperature, pressure, and velocity fields from sensor measurements is essential for effective communication between physical experiments and models. Due to challenging conditions and spatial limitations, sensor placement in nuclear subsystems is highly constrained. This study develops a data-driven optimized constrained sensor placement algorithm to reconstruct the field of interest within a TRi-structural ISOtropic (TRISO) fuel irradiation experiment, a lumped parameter model of a nuclear fuel test rod, and a steam generator. The optimization process leverages reduced-order models of flow physics to achieve highly accurate full-field reconstructions of responses of interest, quantify noise-induced uncertainty, and identify physically feasible sensor locations. These precise sensor-based reconstructions lay the groundwork for digital twinning of subsystems, ultimately leading to a comprehensive digital twin aggregate of a nuclear power plant.