Record Detail Back
Improving Computational Efficiency Using Brute Force Method In Uncertainty/Sensitivity Analysis
- Share to:
IMPROVING COMPUTATIONAL EFFICIENCY USING BRUTE FORCE METHOD IN UNCERTAINTY/SENSITIVITY ANALYSIS. In recent years, probabilistic safety assessment (PSA) has acquired increasing importance. Considering the general and specific validity of the assessment, the uncertainty study is the most challenging issues for PSA, because of the intensive computational demand for assessing the impact of probabilistic variations. Because of its simplicity, the most implemented method in uncertainty study is the Monte Carlo Technique. This technique is the most straightforward and powerful in uncertainty propagation analysis, although computationally inefficient. To improve efficiency is proposed to use the Brute Force method for the sensitivity analysis and the construction of a reduced model. Further, the Monte Carlo simulation is performed to quantify the impact of uncertainties in random variables on the uncertainty in outputs using the obtained reduced model. The proposed technique is then implemented to an analytic loss of cooling accident (LOCA) model. The LOCA model used is a much simplified one, and it is only for test purposed, to proved the usefulness and efficiency of the implemented technique. From this study, it is deduced that using the combination of the response surface method, Monte Carlo simulation with Brute-Force technique, and maximum entropy method, can derived the most important parameter needed for an uncertainty and sensitivity analysis with a minimum computational effort, hence, it is very useful for performing the efficient uncertainty studies.
APA Citation
M. Hendayun. (2007).Improving Computational Efficiency Using Brute Force Method In Uncertainty/Sensitivity Analysis.(Electronic Thesis or Dissertation). Retrieved from https://localhost/etd