Robust design with imprecise random variables and its application in hydrokinetic turbine optimization

Abstract

In robust design, uncertainty is commonly modelled with precise probability distributions. In reality, the distribution types and distribution parameters may not always be available owing to limited data. This research develops a robust design methodology to accommodate the mixture of both precise and imprecise random variables. By incorporating the Taguchi quality loss function and the minimax regret criterion, the methodology mitigates the effects of not only uncertain parameters but also uncertainties in the models of the uncertain parameters. Hydrokinetic turbine systems are a relatively new alternative energy technology, and both precise and imprecise random variables exist in the design of such systems. The developed methodology is applied to the robust design optimization of a hydrokinetic turbine system. The results demonstrate the effectiveness of the proposed methodology.

Keywords:

• robust design
• uncertainty
• optimization
• hydrokinetic energy

Acknowledgements

The authors gratefully acknowledge the support from the Office of Naval Research through contract ONR N000141010923 (Program Manager Dr. Michele Anderson) and the Intelligent Systems Center at the Missouri University of Science and Technology.