Within the Nuclear Safety Analysis Team, my research looks into studying and developing uncertainty quantification methods towards:
- Probabilistic Safety Assessment of nuclear reactors under limited data (the primary focus);
- Statistical Verification and Validation of physics-based models for the prediction of nuclear reactor phenomenon or material properties; and
- Risk analysis of critical sub-systems of a nuclear reactor under limited reliability information.
My work currently relies heavily on machine-learning and numerical techniques including: Bayesian inference, Artificial Neural Network modelling, Principal Component Analysis, and Markov-chain Monte Carlo methods. The areas of topical interest include: Stochastic model updating, Imprecise probabilities, Probability bounds analysis, and Reliability engineering. Besides research, I am also keen on science communication and making science as accessible to as many as possible – an important endeavour considering the significant degree of epistemic uncertainty over nuclear energy and its risks, especially among the general public. Prior to joining SNRSI, I undertook my PhD studies in Risk and Uncertainty quantification (with engineering applications) at the University of Liverpool where the research focus was on developing Markov chain Monte Carlo methods for Stochastic model updating and parameter identification towards structural health monitoring, risk analysis, and stochastic sensitivity analysis.
More information to our work can be found in my website: https://sites.google.com/view/adolphus-lye/