Short-term power prediction of photovoltaic power stations based on Kepler optimization algorithm and VMD-CNN-LSTM...

Area of Science:

• Renewable Energy Systems
• Artificial Intelligence in Energy
• Power Systems Engineering

Background:

• Photovoltaic (PV) power generation is inherently intermittent and fluctuating, posing challenges for stable power system operation.
• Accurate short-term power prediction is crucial for mitigating these fluctuations and ensuring grid reliability.
• Traditional prediction models often struggle to capture the complex nonlinear dynamics of PV power output.

Purpose of the Study:

• To develop an accurate and reliable method for short-term power prediction of photovoltaic power stations.
• To address the challenges of intermittency and fluctuation in PV power generation.
• To improve the overall stability and operational efficiency of the power system.

Main Methods:

• A hybrid model combining Variational Mode Decomposition (VMD), Convolutional Neural Network (CNN), and Long Short-Term Memory (LSTM) networks was constructed.
• VMD was employed to decompose the PV power sequence into intrinsic mode functions, reducing data complexity.
• CNN extracted spatial features from decomposed components, while LSTM captured temporal dependencies, synergistically enhancing prediction.
• The Kepler optimization algorithm (KOA) was integrated to optimize the VMD-CNN-LSTM model parameters and improve prediction accuracy.

Main Results:

• The proposed VMD-CNN-LSTM model integrated with KOA demonstrated significant improvements in prediction accuracy compared to traditional methods.
• Evaluation metrics, including Root Mean Square Error (RMSE) and Mean Absolute Error (MAE), showed substantial optimization.
• The model effectively captured nonlinear relationships and dynamic trends in PV power generation data.

Conclusions:

• The developed Kepler-optimized VMD-CNN-LSTM model offers an innovative and effective solution for short-term PV power prediction.
• This approach significantly enhances prediction accuracy, contributing to the stable operation of power systems with high PV penetration.
• The study provides a valuable methodology for advancing grid integration of renewable energy sources and optimizing power system management.