Abstract:Accident diagnosis is critical for ensuring the safety and reliability of nuclear power plants (NPPs). However, the scarcity of real-world data poses a significant challenge in the development of accurate, and robust, artificial intelligence (AI)-based diagnostic systems. This paper presents a novel approach for addressing this domain issue by leveraging knowledge-distillation techniques. We propose training a lightweight-student model that uses distilled knowledge from multiple teacher models, each of which captures different aspects of the problem, such as noise and time trends. Teacher models and student models were trained on CNS simulation data, and both models were evaluated on real-world proxy data(PCTRAN Simulation data). Through a series of experiments, we demonstrated the effectiveness of our approach in enhancing the performance and robustness of student models. The student model showed improvements in untrained target domain, when knowledge obtained from noisy, time-trend-information containing, and naive teacher models was used instead of a single teacher model. The incorporation of diverse knowledge from different teacher models aids the student model in handling noisy data, capturing temporal dependencies, and generalizing unseen scenarios better. Moreover, our analysis of the relative training time highlighted the computational benefits of using a lighter student model. The student model required significantly less training time than the teacher model, making it more efficient for deployment and online updation in real-world scenarios. This computational efficiency is particularly valuable in the context of NPPs where timely decision making and resource constraints are critical factors. Our findings will contribute to the development of effective, robust, and computationally efficient accident-diagnosis systems for NPPs. By bridging the gap between simulations and real-world scenarios through knowledge distillation, this study paves the way for further advancements in enhancing the safety, reliability, and efficiency of the nuclear power industry.

Application of Bayesian Classifiers for the Accident Diagnosis in Nuclear Power Plants

Combinatorial Techniques for Fault Diagnosis in Nuclear Power Plants Based on Bayesian Neural Network and Simplified Bayesian Network-Artificial Neural Network

Diagnosis of Operational Failures and On-Demand Failures in Nuclear Power Plants: an Approach Based on Dynamic Bayesian Networks

A Method Of Fault Diagnosis Based On Pca And Bayes Classification

Intelligent multi-severity nuclear accident identification under transferable operation conditions

Intelligent Modeling for Nuclear Power Plant Accident Management

Fault prediction method for nuclear power machinery based on Bayesian PPCA recurrent neural network model

AN ACCIDENT DIAGNOSIS METHOD OF HTR-10 BASED ON BAYESIAN INFERENCE MODEL

Fault Diagnosis of Nuclear Power Plants Based on 1D-CNN with Dual Attention Mechanism

Technical research on rapid source term calculation and prediction technology of nuclear accidents based on Bayesian network

Enhancing accident diagnosis in nuclear power plants through knowledge Distillation: Bridging the gap between simulation and Real-World scenarios

Bayesian Long Short-Term Memory Model for Fault Early Warning of Nuclear Power Turbine

Fault Diagnosis Techniques for Nuclear Power Plants: A Review from the Artificial Intelligence Perspective

Pilot Study of Dynamic Bayesian Networks Approach for Fault Diagnostics and Accident Progression Prediction in Htr-Pm

Application of Bayesian Networks in Multi-Hazard Safety Assessment of Nuclear Power Plants

Event diagnosis method for a nuclear power plant using meta-learning

A new sensor fault diagnosis method for gas leakage monitoring based on the naive Bayes and probabilistic neural network classifier

Fault Diagnosis of Nuclear Power Plant Based on Sparrow Search Algorithm Optimized CNN-LSTM Neural Network

Application of genetic algorithms to fault diagnosis in nuclear power plants

Establishing operator trust in machine learning for enhanced reliability and safety in nuclear Power Plants

Rapid source term prediction in nuclear power plant accidents based on dynamic Bayesian networks and probabilistic risk assessment