Abstract
Conference Title: 2018 IEEE International Conference on Electro/Information Technology (EIT) Conference Start Date: 2018, May 3 Conference End Date: 2018, May 5 Conference Location: Rochester, MI, USA Most solar power stations contain hundreds, even thousands, of photovoltaic (PV) modules. Monitoring a solar power station and diagnosing faults in real time are a primary challenge in maintaining the normal operation. A traditional fault detection process is cumbersome and time consuming. In this paper, we are applying a hybrid method for fault detection and localization in serial-parallel configuration using a network of current-voltage sensors-based framework to detect and localize open-circuit, short-circuit, and hotspot faults. Data-analysis is used through a classifier to allow for better fault diagnosis such as the recognition of the environmental factors within a hotspot fault class. Also, we propose a new network model for the strategic placement of voltage sensors that will allow the network to be scaled up to any number of PV modules with 100% accuracy. To validate the proposed algorithm, simulations were developed using MATLAB/Simulink Software Tools. Results under varying conditions are presented.