Integrating Solar Irradiance and Environmental Factors for Predictive Health Assessment of Lead-Acid Batteries

Abstract

Lead-acid batteries remain widely used in solar energy systems due to their affordability and robustness; however, their health and performance are highly sensitive to environmental conditions such as solar irradiance, temperature, and humidity. The degradation mechanisms of these batteries are strongly influenced by fluctuating climatic parameters, necessitating predictive models for effective monitoring. This study aims to develop an integrated prediction framework for lead-acid battery health by correlating solar and environmental variables with electrical performance metrics. Empirical regression models and logistic regression classifiers were constructed using real-time data on light intensity, panel temperature, ambient temperature, humidity, and battery voltage/current. Strong positive correlations were observed between light intensity and both panel temperature (r = 0.95) and DC voltage (r = 0.78), while humidity exhibited negative correlations (r = -0.42 to -0.47), indicating its adverse influence on charging efficiency. The logistic regression model achieved 96% overall classification accuracy with precision and recall above 0.95, effectively predicting healthy battery states though limited by degradation data imbalance. The proposed predictive model demonstrates high reliability in assessing battery health under varying solar conditions, providing a quantitative foundation for optimizing charging cycles, extending battery lifespan, and enabling scalable IoT-based health monitoring in solar energy storage systems.