Composite Imagery-Based Non-Uniform Illumination Sensing for System Health Monitoring of Solar Power Plants

Abstract

Incidence of Non-uniform illumination degrades the affected solar photovoltaic modules temporarily. However, in worst case it may also lead to permanent degradation. Thus, non-uniform illumination sensing in terms of encompassed area over solar photovoltaic module become important for sustainable power generation. In this work, we propose, an integrated framework by combining optical and thermal images for Non-uniform illumination sensing and classification (static and dynamic). The proposed technique detects hotspots along with identification of the nature of shading on the solar photovoltaic modules. Additionally, Hungarian Kalman Filter is implemented for estimating the coverage of the Non-uniform illumination-affected region including its abrupt shape. The proposed estimation technique also calculates the total loss in the output energy of the solar photovoltaic system due to non-uniform illumination. Overall, the proposed methodology develops a hybrid imagery-dependent advanced early warning system for large-scale solar power plants that are cost-effective and bypass multi-sensor data fusion to attain real-time application.

Detection and analysis of deteriorated areas in solar PV modules using unsupervised sensing algorithms and 3D augmented reality

“…Other factors that can affect the performance of PV modules include module soiling, shading, degradation of the anti-reflective coating, and hotspots caused by localized heating. Each of these issues can impact the overall efficiency of the SP station and lead to suboptimal energy generation [17] , [18] .…”