Doctor of Philosophy, The Ohio State University, 2012, Nuclear Engineering

Utilities operating nuclear power plants in the United States are required to demonstrate that their plants comply with the safety requirements set by the U.S. Nuclear Regulatory Commission (NRC). How to show adherence to these limits through the use of computer code surrogates is not always straightforward, and different techniques have been proposed and approved by the regulator. The issue of compliance with regulatory limits is examined by rephrasing the problem in terms of hypothesis testing. By using this more rigorous framework, guidance is proposed to choose techniques to increase the probability of arriving at the correct conclusion of the analysis. The findings of this study show that the most straightforward way to achieve this goal is to reduce the variance of the output result of the computer code experiments. By analyzing different variance reduction techniques, and different methods of satisfying the NRC's requirements, recommendations can be made about the best-practices, that would result in a more accurate and precise result. This study began with an investigation into the point estimate of the 0.95-quantile using traditional sampling methods, and new orthogonal designs. From there, new work on how to establish confidence intervals for the outputs of experiments designed using variance reduction techniques was compared to current, regulator-approved methods. Lastly, a more direct interpretation of the regulator's probability requirement was used, and confidence intervals were established for the probability of exceeding a safety limit. From there, efforts were made at combining methods, in order to take advantage of positive aspects of different techniques. The results of this analysis show that these variance reduction techniques can provide a more accurate and precise result compared to current methods. This means an increased probability of arriving at the correct conclusion, and a more accurate characterization of the risk associated with even (open full item for complete abstract)

Committee: Tunc Aldemir PhD (Advisor); Richard Denning PhD (Committee Member); Marvin Nakayama PhD (Committee Member); Alper Yilmaz PhD (Committee Member) Subjects: Nuclear Engineering; Statistics

Doctor of Philosophy in Engineering, University of Toledo, 2012, College of Engineering

The probabilistic reliability evaluation of power systems is a complex and highly dimensional problem that often requires a large amount of computational resources, particularly processing power and time. The complexity of this problem is only increasing with the advent of the smart grid and its accompanying technologies, such as plug-in hybrid electric vehicles (PHEVs). Such technologies, while they add convenience, intelligence, and reduce environmental impacts, also add dynamic and stochastic loads that challenge the current reliability and security of the power grid. One method that is often used to evaluate the reliability of power systems is Monte Carlo simulation (MCS). As the complexity and dimensionality of a power system grows, MCS requires more and more resources leading to longer computational times. Multiple methods have previously been developed that aid in reducing the computational resources necessary for MCS in order to achieve a more efficient and timely convergence while continuing to accurately assess the reliability of a given system. Examples include analytical state space decomposition, population based metaheuristic algorithms (PBMs), and the use of high performance computing (HPC). In order to address these issues, this dissertation is focused on improving the performance of algorithms used to examine the level of reliability in composite power systems through the use of computational intelligence (CI) and HPC, while also investigating the impact of PHEVs on the power grid at the composite and distribution levels. Contributions include the development and exploration of 3 variations of a new, hybrid algorithm called intelligent state space pruning (ISSP) that combines PBMs with non-sequential MCS in order to intelligently decompose, or prune, a given state space and improve computational efficiency, an evaluation of the use of latin hypercube sampling and low discrepancy sequences in place of MCS, the use of serial and parallel support vecto (open full item for complete abstract)

Committee: Lingfeng Wang Ph.D. (Committee Chair); Mansoor Alam Ph.D. (Committee Co-Chair); Jackson Carvalho Ph.D. (Committee Member); Vijay Devabhaktuni Ph.D. (Committee Member); Mohsin Jamali Ph.D. (Committee Member); Weiqing Sun Ph.D. (Committee Member) Subjects: Artificial Intelligence; Computer Science; Electrical Engineering