International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 05 | May 2025
p-ISSN: 2395-0072
www.irjet.net
Solar Power Generation Prediction using Machine Learning Model Sowmith Reddy1, Piyush Vijay2, Chinmai3, Disha4, Gunottham5 1 2 3 4 5 U.G Student, Dept. of Information Science Engineering, Dayananda Sagar College of Engineering, Bengaluru,
Karnataka, India. ---------------------------------------------------------------------***--------------------------------------------------------------------Gradient Boost Regression model to enhance forecast Abstract- Accurate forecasting of solar energy
accuracy. The model leverages weather data to predict solar energy generation. Results show improved shortterm predictions over traditional methods. The study highlights the effectiveness of machine learning in solar forecasting. Singhal et al. (2022) This study focuses on improving solar power prediction accuracy using weather forecast data. The authors introduce a machine learningbased framework, "Solar-Cast," for precise forecasting. Various algorithms, including ensemble models, are explored for better performance. The results indicate significant accuracy improvements in solar energy prediction. The approach is validated on real-world datasets. Fraccanabbia et al. (2020) The paper tackles the issue of unreliable solar power forecasting due to weather variability. The authors employ ensemble learning techniques to enhance predictive accuracy. The proposed model integrates multiple machine learning methods to improve robustness. Experimental results show superior performance compared to individual models. The study highlights ensemble learning’s potential in renewable energy forecasting. Rusina et al. (2023) The study examines day-ahead solar power forecasting for Mongolia’s power system. An ensemble machine learning approach is developed to improve accuracy. The model combines different regression techniques for better performance. Results demonstrate enhanced forecasting precision compared to conventional models. The approach is applicable in power grid planning. Abdellatif et al. (2022) This paper focuses on forecasting solar photovoltaic (PV) power output using ensemble learning. Various machine learning methods are compared for polycrystalline panelbased predictions. The proposed model effectively captures solar power variations. Results show reduced forecast errors compared to traditional approaches. The study provides insights into optimizing PV power predictions. Chen & Koprinska (2020) The research explores the application of ensemble methods for solar power forecasting. The study integrates different machine learning models to improve predictive performance. The ensemble model achieves better accuracy compared to standalone models. Results highlight the effectiveness of combining diverse algorithms. The approach enhances solar power reliability for grid management. Nayak & Heistrene (2020) The paper addresses inaccuracies in solar power prediction using hybrid models. A hybrid machine learning approach combining multiple techniques is proposed. The model enhances accuracy by leveraging different predictive strengths. Performance evaluations
generation is crucial for the success of large-scale renewable energy facilities, as it depends on evolving weather conditions. This study presents a hybrid model that leverages both machine learning and statistical techniques to predict solar power output. By integrating climate principles with ML algorithms, the proposed approach enhances forecasting accuracy. Additionally, the system incorporates NodeMCU for IoT-based communication, along with components such as an LDR (Light Dependent Resistor), rain sensor, alarm controller, and solar panel for energy generation. This research focuses on implementing ML models using real weather data to optimize the efficiency and reliability of solar energy utilization.
Keywords: Solar energy prediction, hybrid model, machine learning algorithms, climate concepts, IoT communication.
1
INTRODUCTION
Solar energy is a essential renewable energy source that contributes significantly to the global energy mix. However, distance due to weather conditions is a challenge for integration into the existing power grid. Accurate forecasting of solar power generation can lead to better grid management and more efficient use of energy. This research focuses on developing a forecasting system using ML algorithms that uses real weather data received from sensors to forecast solar energy production. By integrating IoT devices such as NodeMCU with sensors and rain sensors such as LDR, the system aims to create a reliable and accurate forecast model. Machine learning approaches have become more popular in recent decades in many companies where data-related challenges are common. Machine learning covers many areas, including data mining, optimization and artificial intelligence, to name a few of the more popular ones. Machine learning approaches attempt to find relationships between input data and output data, whether or not they use mathematical models. After training with the training database, the predictive input data can be fed into a welltrained machine learning model, which can then make predictions.
2
LİTERATURE SURVEY
Tiwari et al. (2018) The paper addresses the challenge of short-term solar irradiance forecasting using Numerical Weather Prediction (NWP). The authors propose a
© 2025, IRJET
|
Impact Factor value: 8.315
|
ISO 9001:2008 Certified Journal
|
Page 1779