Solar Panel Degradation Prediction using Machine Learning: A Comprehensive Approach DOI Creative Commons

Deepanshu Deepanshu,

Kartik Garg,

Harshit Mittal

et al.

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: March 26, 2025

Abstract Solar photovoltaic (PV) systems are central to the world's movement toward renewable power, but their performance declines with time owing a combination of environmental expo- sure and usage stress. In this research, we suggest hybrid machine learning system that incorporates multi-source data such as device logs, weather history, customer endpoints, network endpoints in order make precise predictions about solar panel degradation. The data, which was obtained from London Datastore recorded by UK Power Networks for 480 days, is processed obtain significant features capturing electrical well conditions. High-level feature extraction methods were used stress measures like temperature stress, humidity exposure, voltage drop current total harmonic distortion (THD) Fifteen regression models trained compared based on mean absolute error (MAE), squared (MSE), root (RMSE), coefficient determination (R2) [1, 2]. Our top-performing ensemble, built stacking an artificial neural (ANN), XGBoost, Random Forest, R2 value above 0.96. These findings highlight effectiveness combining various sources advanced engineering proactive maintenance enhanced operational efficiency PV systems.

Language: Английский

Solar Panel Degradation Prediction using Machine Learning: A Comprehensive Approach DOI Creative Commons

Deepanshu Deepanshu,

Kartik Garg,

Harshit Mittal

et al.

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: March 26, 2025

Abstract Solar photovoltaic (PV) systems are central to the world's movement toward renewable power, but their performance declines with time owing a combination of environmental expo- sure and usage stress. In this research, we suggest hybrid machine learning system that incorporates multi-source data such as device logs, weather history, customer endpoints, network endpoints in order make precise predictions about solar panel degradation. The data, which was obtained from London Datastore recorded by UK Power Networks for 480 days, is processed obtain significant features capturing electrical well conditions. High-level feature extraction methods were used stress measures like temperature stress, humidity exposure, voltage drop current total harmonic distortion (THD) Fifteen regression models trained compared based on mean absolute error (MAE), squared (MSE), root (RMSE), coefficient determination (R2) [1, 2]. Our top-performing ensemble, built stacking an artificial neural (ANN), XGBoost, Random Forest, R2 value above 0.96. These findings highlight effectiveness combining various sources advanced engineering proactive maintenance enhanced operational efficiency PV systems.

Language: Английский

Citations

0