It was conducted at nine o’clock in the morning on Monday, 3/18/2024, in the Degla Hall in the College of Engineering / Al-Nahrain University, to discuss the thesis of the master’s student (Hala Sabah Mutashar) in the Department of Electronic and Communications Engineering, which was entitled:

Detecting photovoltaic cell faults using machine learning

This study focuses on enhancing fault detection and classification in stand-alone and grid-connected PV systems by adopting fast and accurate machine learning methods. The research uses MATLAB to simulate 0.25 MW photovoltaic systems in standalone and grid-connected cases. This simulation produces various electrical features, including minimum, maximum, average and range values of currents, voltages and powers. In addition, environmental features and properties such as temperature and radiation are extracted. Specific feature extraction is applied to many state-of-the-art classifiers with appropriate proportions of cross-validation (CV) for error detection and type classification. The four classification categories are represented by normal operation (fault-free condition) and electrical faults: string-to-string fault (F1), string-to-string fault (F2), and string-to-ground fault (F3). The study uses simple data preprocessing to identify features with missing values in the data set, performs an appropriate imputation to find the missing values based on the KNN algorithm, and then applies minimum and maximum data normalization.

The discussion committee consisted of the following individuals, their names and their affiliations: