An Algorithm for Filtering the Time-series I-V curves of a PV plant

This paper presents a Machine learning-based algorithm to filter the time-series I-V curves collected from a PV plant. The filter’s objective is detecting and segregating the normal looking I-V curves from the abnormal ones. A non-linear Machine learning regression model was used for this purpose. Initially the model is trained on the labeled data and then tested on a unknown data. The proposed model is trained and tested on the time-series I-V curves of experimental PV strings installed in the field. The obtained model coefficients, results are discussed in the paper.