In recent years, grid connected renewable systems are increasingly becoming part of the distributed power generation. In this work, a two-stage power conversion topology consisting of step-up dc-dc converter followed by three-phase inverter is proposed and investigated for grid connect applications from low voltage PV/fuel cell sources. Different formats of winding layout are explored to determine the optimum transformer layout considering the performance of dc-dc converter specifically of high voltage ratio in this work. The impact of winding configuration on the leakage inductance and winding capacitance is also studied in detail.

As power levels increase, reduction in switching frequencies is crucial to maintain efficiencies. Based on a detailed study of inverter efficiencies, it is shown that a switching pattern based on Selective Harmonic Elimination (SHE) could be advantageous in such grid connected operation. Solutions based on convex and Sequential Quadratic Programming method for SHE switching pattern determination that can be to solve the transcendental equations are discussed. In this work, a novel approach to design the LCL filter based on a cost-per-joule minimization is proposed and compared with other methods presented in literature. The approach also uses weightage factors for accounting for the difference in costs between the inductors and also with the capacitor. It is seen that the approach results in considerably lower values for filter inductances. The impact of the lower filter components on inverter device selection and costs is also explored.

A suitable control algorithm utilizing a fixed switching pulse pattern in a grid connected mode is proposed. Approaches for synchronization, active and reactive power injection are explored fur-ther. The control system is designed which then achieves explicit control of voltage magnitude and phase angle (with respect to grid voltage) as a means of achieving reactive and active power injection respectively, independently. Simulation studies consider a 10 kVA system, while the laboratory prototype is a 1 kVA unit. In both cases a dc source of 48 V is considered. Finally, the proposed design is experimentally validated through the laboratory prototype. The experimental results shown for different cases are seen to validate the method proposed and the simulation studies.