International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 11 | Nov 2025
p-ISSN: 2395-0072
www.irjet.net
Solar Energy Optimization Using MPPT and AI-Based Prediction Mrs. Kalpana S1, Ajay Kumar S2, Divya M3, Karthik G.D4, Usha T5. 1 Assistant Professor Department of Electrical and Electronics Engineering
PES Institute of Technology and Management, Shivamogga, Karnataka, India 2,3,4,5 BE Final Year Students, Department of Electrical and Electronics Engineering
PES Institute of Technology and Management, Shivamogga, Karnataka, India -------------------------------------------------------------------------***-----------------------------------------------------------------------significant energy losses in these solar installations. Abstract - Unpredictable changes in weather conditions and
Appropriate maximum power point tracking technology implementation in these systems could provide energy capture improvements of 15-30%. Finally, the lack of realtime system monitoring and remote access restricts maintainability and user engagement, which in turn results in extended periods of undetected performance loss. Managing solar energy today requires the use of smart systems which can determine on the fly optimal configurations and predictive analytics that help maximize energy capture and Software and Communication Infrastructure: lower operational costs. The use of IoT devices and advanced control automation provides new and powerful means of delivering solar energy systems automation which self-adjust to prevailing weather changes and provide users with complete operational feedback. This research addresses the identified cutting-edge gaps with the introduction of a novel dual-controller system design that combines hardware-based MPPT optimization with IoT enabled extensive monitoring and cloud control. The system leverages the real-time power of the Arduino Mega 2560 for responsive power optimization and employs the processing power, connectivity, and control of the Raspberry Pi for intelligent system data, external API integration, and user interface design.
poorly designed solar power system control mechanisms typically lead to under-performance in solar photovoltaic systems. The research presented here goes one step further in developing a more integrated approach to optimizing solar energy systems which focuses on the combination of real-time photovoltaic systems Maximum Power Point Tracking, comprehensive IoT (Internet of Things) monitoring, and artificial intelligence (AI) forecasting. The system designed for this purpose incorporates an Arduino Mega 2560 microcontroller that runs a Perturb and Observe (P And O) power optimizing algorithm and a Raspberry Pi that serves as the intelligent data gateway for sensor data. The Raspberry Pi also connects to the NASA POWER API to download weather data and provides access to a cloud dashboard for users. The testing conducted demonstrated a 15 25%improvement on energy capture compared to standard solar systems. The monitored system provides users engagement on the maintenance of the system. This work provides a comprehensive solution for the gap between traditional solar systems to a forward intelligent energy management system. This approach is economically feasible for distributed photovoltaic systems.
Key Words: Solar photovoltaic systems, Maximum Power Point Tracking, Perturb and Observe algorithm, Internet of Things, Arduino microcontroller, Raspberry Pi, cloud monitoring, buck-boost converter.
2. SYSTEM ARCHITECTURE AND DESIGN A. Overall System Configuration
1. INTRODUCTION Advances in the adoption of sustainable energy technologies make solar photovoltaics a core component of any modern renewable energy system. Yet, there are considerable challenges to the variability of these systems and their sensitivity to changing environmental conditions. A solar panel’s performance is influenced to a considerable degree, if not the most, by non-linear, voltage-current characteristics which change continuously with the intensity of solar radiation, ambient temperature, and partial shading. These complex conditions can lead to a significant loss in the amount of energy that is extracted under conventional methods that use fixed parameters. Passive mode operation is the default for most solar installations. These systems convert the solar energy that is available and do not take dynamic optimization steps. This passive approach is responsible for
© 2025, IRJET
|
Impact Factor value: 8.315
Fig. 1. System Architecture of the solar energy optimization system. The solar energy optimization system proposed includes a creatively designed two-controller construction which is specifically designed for optimizing real-time performance
|
ISO 9001:2008 Certified Journal
|
Page 884