Real-Time Internet of Things Predictive Analytics for On-Grid Solar Power and Efficiency in Tropical Smart Cities

Abstrak

The performance of On-Grid Solar Power Plants is strongly influenced by environmental conditions, particularly solar irradiance and module temperature. Therefore, an accurate prediction model is required to evaluate system performance under actual operating conditions. Previous studies have generally relied on historical meteorological datasets or simulation-based approaches and have rarely integrated real-time Internet of Things (IoT) monitoring with PVsyst simulation, Simulink modeling, and Multiple Linear Regression (MLR) within a unified predictive framework. This study aims to develop a predictive model for the output power and efficiency of a 160.05 kWp On-Grid Solar Power Plant installed at the Integrated Waste Processing Facility, Sepaku District, North Penajam Paser Regency, East Kalimantan. The novelty of this research lies in the integration of real-time IoT monitoring, PVsyst simulation, Simulink modeling, and MLR to estimate photovoltaic output power and system efficiency under actual operating conditions. The proposed model integrates real-time IoT-based monitoring, PVsyst simulation, Simulink modeling, and MLR. Real-time operational data, including solar irradiance and module temperature, were used to develop and validate the prediction model. The PVsyst simulation estimated an annual energy production of 58,418.25 kWh with a Performance Ratio (PR) of 78.30%. Model validation produced a Mean Absolute Percentage Error (MAPE) of 3.15% for output power prediction and 2.38% for system efficiency prediction, indicating high prediction accuracy. The developed model can accurately predict the output power and efficiency of the On-Grid Solar Power Plant and can be used to support performance evaluation based on actual operational data.

Identifier

• OpenAlex: https://openalex.org/W7201860939
• DOI: https://doi.org/10.61132/manufaktur.v4i2.1514