Spatio-Temporal Graph Learning for Large-scale Photovoltaic Degradation Patterns Analysis

Yangxin Fan, Raymond Wieser, Laura Bruckman, Yinghui Wu, Roger French Case Western Reserve University, Cleveland, OH, United States

Photovoltaic (PV) power stations have become an integral component to the global sustainable energy production landscape. Accurately estimating the performance of PV systems is critical to their feasibility as a power generation technology and as a financial asset. One of the most challenging problems in assessing the Levelized Cost of Energy (LCOE) of a PV system application is to understand and estimate the long-term Performance Loss Rate (PLR) for large fleets of PV inverters. We propose a novel Spatio-Temporal Dynamic Graph Neural Network (st-DynGNN) to perform fleet-wide PLR estimation. st-DynGNN integrates spatio-temporal coherence and graph attention to separate PLR as a long-term “aging” trend from multiple fluctuation terms in the PV input data. st-DynGNN imposes flatness and smoothness regularization to ensure the disentanglement between aging and fluctuation. We have evaluated st-DynGNN on three large-scale Photovoltaics (PV) datasets spanning a time period of 10 years. Our results show that st-DynGNN reduces Mean Absolute Percent Error (MAPE) and Euclidean Distances by 34.74% and 33.66% compared to the state-of-the-art methods.