Midterm Power Load Forecasting Model Based on Kernel Principal Component Analysis and Back Propagation Neural Network...

Area of Science:

• Electrical Engineering
• Artificial Intelligence
• Data Science

Background:

• Accurate power load forecasting is crucial for efficient energy management and grid stability.
• Traditional forecasting methods often struggle with high-dimensional and complex load data.
• Midterm forecasting requires robust models capable of handling temporal dependencies and non-linear patterns.

Purpose of the Study:

• To develop and validate a novel hybrid model for improving midterm power load forecasting accuracy.
• To leverage dimensionality reduction techniques for enhanced neural network performance.
• To provide a reliable forecasting tool for daily peak load prediction.

Main Methods:

• Kernel Principal Component Analysis (KPCA) for input space dimensionality reduction.
• Back Propagation Neural Network (BPNN) for load pattern recognition.
• Particle Swarm Optimization (PSO) for optimizing neural network parameters.
• A hybrid approach combining KPCA, BPNN, and PSO for midterm load forecasting.

Main Results:

• The proposed model achieved a mean absolute percent error of 1.39% on European power load data.
• KPCA effectively reduced data dimensionality, improving model efficiency.
• The hybrid model demonstrated superior accuracy compared to traditional methods.

Conclusions:

• The combined KPCA and BPNN model offers a feasible and valid solution for accurate midterm power load forecasting.
• The integration of PSO further optimizes the forecasting performance.
• This approach provides a significant advancement in intelligent energy management systems.