Photovoltaic system, control, analysis and fault diagnosis

Abstract

This work deals with fault diagnosis using machine learning algorithms of the inverter used in photovoltaic systems that supply an insulated electrical load and how to safely transfer the current to devices. This thesis discusses the multicellular inverter and describes how this is affected in cases of faults on the load current. And use two modes of control In order to compare in terms of functionality under failures, load current save and Smoothest, and in terms of accuracy built classification model To use sliding mode control mode and exact linearization mode, this is for Purpose of comparison in terms of system performance during failure And the extent of its impact on the load current by examining the shape of its signal And the robustness analysis of the two controls was not significantly affected by defects and their explanation.