This thesis addresses the design and implementation of micro-inverters for grid-connected single-phase photovoltaic (PV) systems. Despite the existing research issues concerning Micro-inverters, they have recently become very attractive due to their modularity and capability of independent maximum power point tracking (MPPT). The complexity in the design of micro-inverters stems from strict grid connection standards and high expectations of compactness, large amplification gain, high efficiency over a wide range of operating conditions and excellent output power quality. Moreover, since micro-inverters are exposed to a wide temperature range, the reliability and life-time of this technology are major problems. The main limiting factor in the life-time of micro-inverters is the use of large electrolytic capacitors for power decoupling.

New circuit configuration and control structures to design a compact and efficient micro-inverter with high quality and robust output power injection capabilities are introduced in this thesis. In the proposed topology electrolytic capacitors are eliminated, removing the obstacles in achieving a durable and reliable design. To achieve a compact design, the proposed micro-inverter consists of a soft-switching high frequency resonant converter at the input and a hard-switching lower frequency inverter with a high order filter at the output.

Small and large signal models of the resonant converter are obtained to design controllers. A new optimal controller and a design method are also proposed for the inverter that yield robust performance with a high quality output in the presence of grid voltage harmonics, impedance uncertainties and frequency changes. Furthermore, using a new nonlinear control strategy, a direct instantaneous power control method is proposed to achieve fast active and reactive power injections into the grid without using the measurement or calculation of active and reactive powers.

A comprehensive steady state analysis is carried out to arrive at a final design that ensures optimum responses for all operating conditions. Moreover, for all proposed controllers, stability analysis is performed to guarantee sufficient stability margins accounting for uncertainties and nonlinearities. Analytical, simulation and experimental results are presented to verify the effectiveness of the proposed methods.