A combination predicting methodology based on T-LSTNet_Markov for short-term wind power prediction

Area of Science:

• Renewable Energy Systems
• Computational Intelligence
• Time Series Forecasting

Background:

• Wind power is crucial for clean energy but faces challenges due to generation volatility.
• Accurate short-term wind power prediction is essential for stable grid integration.
• Existing prediction models struggle with the inherent uncertainty of wind energy.

Purpose of the Study:

• To develop and validate a hybrid model for improved short-term wind power forecasting.
• To address the challenges posed by wind power intermittency in grid-connected systems.
• To enhance the accuracy of wind energy prediction using advanced computational techniques.

Main Methods:

• Data preprocessing and cleaning of raw wind power data.
• Forecasting using the T-LSTNet model on preprocessed data.
• Error correction using k-means++ clustering and a Weighted Markov process.

Main Results:

• The T-LSTNet model provided initial wind power forecasts.
• Error analysis revealed discrepancies between predicted and actual values.
• The combined T-LSTNet_markov model demonstrated superior prediction accuracy after error correction.

Conclusions:

• The proposed combined model effectively improves short-term wind power prediction accuracy.
• Error correction techniques are vital for enhancing the reliability of wind energy forecasts.
• The T-LSTNet_markov approach offers a promising solution for grid-connected wind power systems.