Shallow Recurrent Decoders for State Estimation in Nuclear Reactors
SHRED (Shallow REcurrent Decoder) is a cutting-edge neural network architecture designed to reconstruct high-dimensional spatio-temporal fields from sparse sensor data. By integrating a Long Short-Term Memory (LSTM) network to capture temporal dynamics and a Shallow Decoder Network (SDN) to map these dynamics back to the original high-dimensional space, SHRED provides a robust solution for scenarios with limited or randomly placed sensors (Riva et al., 2024).
🔑 Key Features
⚡ Efficient Sensor Utilization
SHRED can accurately reconstruct complex spatio-temporal dynamics using as few as three sensors, even if they are randomly positioned. This capability reduces the need for large-scale sensor networks and eliminates the need for complex sensor placement optimization.
đź’» Computational Efficiency
Training SHRED on compressed data enables efficient computation, making it possible to run models on standard hardware without requiring extensive hyper-parameter tuning.
🌍 Versatility Across Domains
SHRED has been successfully applied in various fields, including fluid dynamics and plasma physics, proving its adaptability to different types of spatio-temporal data.
🚀 Recent Applications in Nuclear Reactor Monitoring
In collaboration with the University of Washington, we have extended SHRED for nuclear reactor monitoring. Using out-of-core neutron flux measurements, we demonstrated that SHRED can accurately reconstruct the full state vector of a reactor, including multiple coupled field variables. This advancement holds great potential for real-time monitoring and control in nuclear reactors, particularly in environments where in-core sensing is difficult.
SHRED architecture applied to state estimation in nuclear reactors, particularly the **Molten Salt Fast Reactor** (Riva et al., 2024).
SHRED continues to evolve, offering promising solutions for efficient and accurate state estimation in complex systems with limited sensor data.
For a visual overview of SHRED’s application in nuclear reactor monitoring, check out the YouTube Video.
References
2024
arXiv
Robust State Estimation from Partial Out-Core Measurements with Shallow Recurrent Decoder for Nuclear Reactors
Stefano Riva, Carolina Introini, Antonio Cammi, and 1 more author
Sep 2024
preprint available at \hrefhttps://arxiv.org/abs/2409.12550https://arxiv.org/abs/2409.12550
Reliable, real-time state estimation in nuclear reactors is of critical importance for monitoring, control and safety. It further empowers the development of digital twins that are sufficiently accurate for real-world deployment. As nuclear engineering systems are typically characterised by extreme environments, their in-core sensing is a challenging task, even more so in Generation-IV reactor concepts, which feature molten salt or liquid metals as thermal carriers. The emergence of data-driven methods allows for new techniques for accurate and robust estimation of the full state space vector characterising the reactor (mainly composed by neutron fluxes and the thermal-hydraulics fields). These techniques can combine different sources of information, including computational proxy models and local noisy measurements on the system, in order to robustly estimate the state. This work leverages the Shallow Recurrent Decoder (SHRED) architecture to estimate the entire state vector of a reactor from three, out-of-core time-series neutron flux measurements alone. Specifically, the Molten Salt Fast Reactor, in the EVOL geometry (Evaluation and Viability of Liquid Fuel Fast Reactor System project), is demonstrated as a test case, with neutron flux measurements alone allowing for reconstruction of the 20 coupled field variables of the dynamics. This approach can further quantify the uncertainty associated with the state estimation due to its considerably low training cost on compressed data. The accurate reconstruction of every characteristic field in real-time makes this approach suitable for monitoring and control purposes in the framework of a reactor digital twin.
@misc{riva2024robuststateestimationpartial,title={Robust State Estimation from Partial Out-Core Measurements with Shallow Recurrent Decoder for Nuclear Reactors},author={Riva, Stefano and Introini, Carolina and Cammi, Antonio and Kutz, J. Nathan},year={2024},eprint={2409.12550},journal={under review to Progress in Nuclear Energy},archiveprefix={arXiv},month=sep,primaryclass={physics.ins-det},doi={10.48550/arXiv.2409.12550},url={https://arxiv.org/abs/2409.12550},note={preprint available at \href{https://arxiv.org/abs/2409.12550}{https://arxiv.org/abs/2409.12550}}}
