Development of monitoring and automatic fault detection solutions for grid-connected photovoltaic systems

Abstract

In this Final Thesis work, the development of a new monitoring and automatic fault detection system for grid-connected photovoltaic systems is presented and described in its details. This product has been developed in JavaScript and HTLM protocols and it allow real time an online performance monitoring and comparison together with fault detection and causes diagnosis. The presented solution is focus on the DC side of the PV system and it includes: a simple and effective simulation model (based on the calculation normalized energy yield from the environmental irradiance and the module temperature evolution), a real time monitoring and supervision tool (based on the comparison between simulated and actual power production) and an innovative fault detection system (based on the analysis of the miscellaneous capture losses, the fault signal frequency and two indicators). The two indicators used to distinguish among the most common faults are: the ratio between maximum power point current and the short circuit one, and the ratio between maximum power point voltage and the open circuit one. An intuitive and user friendly visual interface has also been created as an on-line web page to allow the power plant operator an easy and prompt supervision of the behavior of multiple PV plants. The developed solution has been successfully test on an actual grid-connected PV system in real and faulty operation conditions.