Analysis, modeling and design of energy management and multisource power systems

Transformative impacts on our energy security rely on creative approaches for consumption and generation of electricity. Technological contributions can impact both areas if they focus on problems of scale. For example, occupancy-based electrical loads (HVAC and lighting) accounted for roughly 50% of the total consumed electricity in the U.S. in 2008. Meanwhile, roughly 50% of consumed oil in the U.S. is imported. The U.S. Department of Energy has appropriately identified “sensing and measurement” as one of the “five fundamental technologies” essential for achieving energy security. Complementing reductions in consumption with increases in deployment of fossil-fuel-independent generation (solar and wind) and energy storage (batteries, capacitors and fuel cells) will yield a two-fold impact. Lofty energy security goals can be made realizable by aggressive application of inexpensive technologies for minimizing waste and by maximizing energy availability from desirable sources. Long-standing problems in energy consumption and generation can be addressed by adding degrees of freedom to sensing and power conversion systems using multiple electrical sources. This principal drove the invention of the hybrid electric vehicle, which achieves efficiency increases by combining the energy capacity of gasoline with the flexible storage capability of batteries. Similarly, fresh strategies for electrical circuit design, control, and estimation in systems with multiple electrical sources can minimize consumption, extend the useful life of storage, and improve the efficiency of generation. A solar array constitutes a grid or network of panels or cells that may best be modeled and treated as independent sources needing careful control to maximize overall power generation. A fuel cell stack, an array of sources in its own right, is best used in a hybrid arrangement with batteries or capacitors to mitigate the impact of electrical transients. Meanwhile, room lighting constitutes a network of multiple electrostatic field sources that can be particularly useful for occupancy detection. Exploiting performance benefits of multi-source electrical networks requires an increased flexibility in the analysis required to make informed design choices. This thesis addresses the added complexity with linear analytical and modeling approaches 3 that reveal the salient features of complicated multisource systems. Examples and prototypes are presented in capacitive sensing occupancy detectors, hybrid power systems and multi-panel solar arrays. Thesis Supervisor: Steven B. Leeb Title: Professor, EECS & ME, MacVicar Faculty Fellow