International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 12 | Dec 2025
p-ISSN: 2395-0072
www.irjet.net
Solar Power Management and SoC Prediction System Using Machine Learning Ansh Shinde1, Neeraj Maggavi1, Akshay Siddannavar1, Siddarood Naragund1 1Student, Dept. of Electronics & Communication, S.G.Balekundri Institute of Technology, Karnataka, India ---------------------------------------------------------------------***---------------------------------------------------------------------1.1 Motivation Abstract - Effective management of energy storage is
crucial for off-grid solar systems. Traditional charge controllers depend on fixed voltage thresholds and often cannot predict critical battery drains before they happen. This paper introduces a smart solar management system that uses a Hybrid IoT Architecture, combining ThingSpeak for dataset creation and Blynk for real-time user monitoring. The system is designed on the ESP32 microcontroller and works with INA219, BME280, and BH1750 sensors. A unique Edge AI method is used where a Multiple Linear Regression (MLR) model, trained on historical data collected through ThingSpeak, is deployed directly on the ESP32 to predict the future State of Charge (SoC). This enables the system to manage loads (relays) proactively and inform users through the Blynk app and a local buzzer. Experimental results indicate that the predictive model accurately estimates battery backup time and significantly reduces unexpected power outages compared to traditional methods.
The main reason for this project is the limitations of current "Voltage-Based" monitoring systems. Voltage is a delayed signal of battery capacity. By the time a typical controller notices low voltage, the battery might be significantly drained. Furthermore, standard controllers cannot see environmental factors. They do not consider how temperature or sunlight intensity impacts the charging rate. There is a clear need for a system that can:
Key Words: IoT, Solar Energy, Machine Learning, SoC Prediction, ESP32, Battery Management, Remote Monitoring, INA219.
Impact Factor value: 8.315
Enable users to remotely monitor and control their energy usage.
Automatically reduce non-essential loads, like heavy sockets, to extend the runtime of critical appliances, like lights.
The main goal of this work is to design a low-cost, predictive energy manager using the ESP32 NodeMCU. Unlike solutions that rely on the cloud, this system uses Edge AI, which means the decision-making algorithm runs locally on the microcontroller.
Solar energy is quickly becoming the best renewable energy source for off-grid use. However, the reliability of a solar photovoltaic (PV) system is mainly tied to its energy storage unit, the battery. In rural and off-grid settings, battery health often suffers because of inconsistent charging cycles caused by unpredictable weather, like cloudy skies or rain. A major challenge for these systems is the absence of smart energy management. Most traditional solar charge controllers available today work in a "reactive" way. They check the battery's terminal voltage and only cut off the load when the voltage drops below a set level, usually 11.5V or 11.7V. This approach often doesn't work well because voltage can change a lot with heavy loads. This can lead to early disconnections or, on the other hand, deep discharges that can seriously harm the battery. Additionally, these systems don’t inform users about the actual "backup time" left, which can leave them unprepared for unexpected power outages. To fill these gaps, this paper introduces a Smart Solar Power Management System that uses Edge Artificial Intelligence (AI) and a Hybrid IoT Architecture. By combining the ESP32 microcontroller with a Multiple Linear Regression (MLR) model, the proposed system turns a regular solar setup into a smart device that can forecast future energy availability.
|
Predict power failure before it happens.
1.2 Objectives
1.INTRODUCTION
© 2025, IRJET
Data Driven: We use ThingSpeak to log historical sensor data, including Voltage, Current, Lux, and Temperature, to train our prediction model. User Centric: We use the Blynk IoT App to provide real-time visualization and manual control over household loads. Predictive Logic: The system estimates the State of Charge (SoC) and actively manages the load relays to prevent blackouts and extend battery life.
2. PROPOSED SYSTEM DESIGN The proposed system is built as a standalone, smart energy manager that runs on a "Sense-Predict-Act" cycle. Unlike passive data loggers, this system takes action to protect the battery using real-time and predicted data. The structure is divided into two main areas: the Hardware Layer and the IoT/Software Layer.
|
ISO 9001:2008 Certified Journal
|
Page 1139