Abstract:
This M.Sc thesis present the high gain DC-DC converter (HGDDC) for solar photovoltaic
system (SPVs) with a multi-functional grid-tied inverter (MFGTI). The HGDDC step up the low
SPVs voltage to a constant high gain DC voltage then to be easily integrated with utility grid
voltage, at meantime to minimize the voltage ripple and reduce the stress of switches at output bus
bar. In addition MFGTI convert the stepped up high gain voltage into demanded three phase utility
grid voltage. While integrating the high gain voltage with utility grid voltage there are challenges
like: grid voltage synchronization and power quality problems that are happened to the system due
to non-linear loads, and SPVs irradiation changes. Therefore, to overcome of the grid
synchronization and power quality issues, such as total current harmonic distortion, reactive power
and to enhance the power factor, the author proposed a developed SRF control topology based on
artificial neural network (ANN) techniques. This topology enable the MFGTI to inject the active
power of SPVs to the utility gird Simulteniously to compensate the reactive power, reduce the total
current harmonic, to enhance the power factor of the system and at main time to maintain the dc
link voltage with very less voltage fluctuation. Eventually, the simulated results validate the
satisfactory working of the proposed control topology under various solar irradiation, and
nonlinear load conditions, it’s worthy to mention by implementing the mention topology the total
current harmonics of the system reduced within IEEE-519 standards. To the end, a comparison
study between ANN controller and dynamic PI of the synchronize reference frame (SRF) control
theory are tested. Several simulations and prototype results are depicted to verify and validate the
influenced of ANN based control.
The proposed model is performed by MATLAB®/Simulink software, and the prototype work
is done through Arduino Atmega 2560 controller.