ABSTRACT
Solar power is becoming ever more popular in a variety of applications. It is particularly
attractive because of its abundance, renewability, and environment friendliness. Solar powered
spacecraft systems have ever-expanding loads with stringent power regulation specifications.
Moreover, they require a light and compact design of their power system. These constraints
make the optimization of power harvest from solar arrays a critical task.
Florida Power Electronics Center (FPEC) at UCF set to develop a modular fault-tolerant
power system architecture for space applications. This architecture provides a number of very
attractive features including Maximum Power Point Tracking (MPPT) and uniform power stress
distribution across the system.
MPPT is a control technique that leads the system to operate its solar sources at the point
where they provide maximum power. This point constantly moves following changes in ambient
operating conditions. A digital controller is setup to locate it in real time while optimizing other
operating parameters. This control scheme can increase the energy yield of the system by up to
45%, and thus significantly reduces the size and weight of the designed system.
The modularity of the system makes it easy to prototype and expand. It boosts its
reliability and allows on-line reconfiguration and maintenance, thus reducing down-time upon
faults.
This thesis targets the analysis and optimization of this architecture. A new modeling
technique is introduced for MPPT in practical environments, and a novel digital power stress
distribution scheme is proposed in order to properly distribute peak and thermal stress and
improve reliability.
iii