Comparison Analysis of Machine Learning Techniques for Photovoltaic Prediction Using Weather Sensor Data.

Area of Science:

• Renewable Energy Systems
• Electrical Engineering
• Environmental Science

Background:

• Solar power is a growing renewable energy source, offering clean electricity.
• Unpredictable weather causes fluctuations in solar power generation, posing challenges for grid management.
• Accurate forecasting is essential for grid operators and energy providers to ensure power continuity and manage energy storage.

Purpose of the Study:

• To propose an efficient framework for comparing solar power forecasting techniques.
• To forecast solar power generation 36 hours in advance.
• To analyze the performance of state-of-the-art forecasting methods for a specific solar power plant.

Main Methods:

• Development of a comparative analysis framework for solar power forecasting.
• Application of the framework to data from the Yeongam solar power plant in South Korea.
• Evaluation of various state-of-the-art forecasting techniques.

Main Results:

• The study presents a comparative analysis of different solar power forecasting methods.
• The results highlight the effectiveness of the proposed framework in evaluating forecasting accuracy.
• Specific findings on the performance of state-of-the-art techniques are detailed.

Conclusions:

• The proposed framework provides an efficient method for comparing solar power forecasting techniques.
• Accurate forecasting is vital for integrating solar energy into the power grid.
• The study contributes to the optimization of solar power generation and grid management.