Engineering Commons^™

Considerations For The Design Optimization Of Floating Offshore Wind Turbine Blades, Evan M. Gaertner

Doctoral Dissertations

Floating offshore wind turbines are an immature technology with relatively high costs and risk associated with deployment. Of the few floating wind turbine prototypes and demonstration projects deployed in real metocean conditions, all have used standard turbines design for onshore or offshore fixed bottom conditions. This neglects the unique unsteady aerodynamics brought on by floating support structure motion. While the floating platform has been designed and optimized for a given rotor, the global system is suboptimal due to the rotor operating in conditions outside of which it was design for. If the potential offered by floating wind turbines is to …

Modeling And Optimization Algorithm For Sic-Based Three-Phase Motor Drive System, Ren Ren

Doctoral Dissertations

More electric aircraft (MEA) and electrified aircraft propulsion (EAP) becomes the important topics in the area of transportation electrifications, expecting remarkable environmental and economic benefits. However, they bring the urgent challenges for the power electronics design since the new power architecture in the electrified aircraft requires many benchmark designs and comparisons. Also, a large number of power electronics converter designs with different specifications and system-level configurations need to be conducted in MEA and EAP, which demands huge design efforts and costs. Moreover, the long debugging and testing process increases the time to market because of gaps between the paper design …

Computational Model For Neural Architecture Search, Ram Deepak Gottapu

Doctoral Dissertations

"A long-standing goal in Deep Learning (DL) research is to design efficient architectures for a given dataset that are both accurate and computationally inexpensive. At present, designing deep learning architectures for a real-world application requires both human expertise and considerable effort as they are either handcrafted by careful experimentation or modified from a handful of existing models. This method is inefficient as the process of architecture design is highly time-consuming and computationally expensive.

The research presents an approach to automate the process of deep learning architecture design through a modeling procedure. In particular, it first introduces a framework that treats …