@article{chen_bao_dinh_2024, title={Evaluating the reliability of machine-learning-based predictions used in nuclear power plant instrumentation and control systems}, volume={250}, ISSN={["1879-0836"]}, DOI={10.1016/j.ress.2024.110266}, abstractNote={The field of data-driven, neural-network-based machine learning (ML) has seen significant growth, with applications in various information and control systems. Despite promising real-world uses, the reliability of models remains questionable. Conventionally, reliability is assessed based on predictive fidelity, accuracy, and training effectiveness; however, quality developmental procedures and excellent training performance metrics do not guarantee operational reliability. Instead, an ML model's predictive performance depends on the training set's representativeness to the intended operational space. It is known that ML algorithms excel at interpolation but struggle with extrapolation tasks. Anomalies and feature drift can also reduce operational performance. Determining whether a new sample is an interpolation or extrapolation task involves out-of-distribution (OOD) detection for assessing its proximity to the existing training data. Thus, we present a real-time, model-agnostic individual prediction reliability evaluation method called Data Auditing for Reliability Evaluation (DARE) for applying OOD detection to the training dataset. We demonstrate on a feedforward neural network ML-integrated digital twin for predicting fuel centerline temperatures during loss-of-flow transients. DARE acts as a "data supervisor" in determining the model's applicability under different operating conditions. In this manner, we demonstrate how training data can serve as inductive evidence to support the reliability of ML predictions.}, journal={RELIABILITY ENGINEERING & SYSTEM SAFETY}, author={Chen, Edward and Bao, Han and Dinh, Nam}, year={2024}, month={Oct} }
 @article{bao_zhang_shorthill_chen_lawrence_2023, title={Quantitative evaluation of common cause failures in high safety-significant safety-related digital instrumentation and control systems in nuclear power plants}, volume={230}, ISSN={["1879-0836"]}, DOI={10.1016/j.ress.2022.108973}, abstractNote={Digital instrumentation and control (DI&C) systems at nuclear power plants (NPPs) have many advantages over analog systems. They are proven to be more reliable, cheaper, and easier to maintain given obsolescence of analog components. However, they also pose new engineering and technical challenges, such as possibility of common cause failures (CCFs) unique to digital systems. This paper proposes a Platform for Risk Assessment of DI&C (PRADIC) that is developed by Idaho National Laboratory (INL). A methodology for evaluation of software CCFs in high safety-significant safety-related DI&C systems of NPPs was developed as part of the framework. The framework integrates three stages of a typical risk assessment—qualitative hazard analysis and quantitative reliability and consequence analyses. The quantified risks compared with respective acceptance criteria provide valuable insights for system architecture alternatives allowing design optimization in terms of risk reduction and cost savings. A comprehensive case study performed to demonstrate the framework's capabilities is documented in this paper. Results show that the PRADIC is a powerful tool capable to identify potential digital-based CCFs, estimate their probabilities, and evaluate their impacts on system and plant safety.}, journal={RELIABILITY ENGINEERING & SYSTEM SAFETY}, author={Bao, Han and Zhang, Hongbin and Shorthill, Tate and Chen, Edward and Lawrence, Svetlana}, year={2023}, month={Feb} }
 @article{iskhakov_dinh_chen_2021, title={Integration of neural networks with numerical solution of PDEs for closure models development}, volume={406}, ISSN={["1873-2429"]}, url={http://www.scopus.com/inward/record.url?eid=2-s2.0-85107023012&partnerID=MN8TOARS}, DOI={10.1016/j.physleta.2021.127456}, abstractNote={Modeling often requires closures to account for the multiscale/multiphysics nature of certain phenomena. Recently, there has been interest in the application of machine learning (ML) for their development. Most of the applications are purely data-driven; however, incorporation of the knowledgebase is an opportunity to enhance flexibility and predictive capability of ML models. This paper presents a PDE-integrated ML framework. PDEs are solved using convolutional operators and integrated with neural networks (NNs). Such integration allows one to train the NNs directly on observed field variables. To demonstrate the framework's viability, NNs are integrated with heat conduction, Navier-Stokes, and RANS equations.}, journal={PHYSICS LETTERS A}, author={Iskhakov, Arsen S. and Dinh, Nam T. and Chen, Edward}, year={2021}, month={Aug} }