A VVWBO-BVO-based GM (1,1) and its parameter optimization by GRA-IGSA integration algorithm for annual power load...

Area of Science:

• Electrical Engineering
• Computational Intelligence
• Forecasting Science

Background:

• Annual power load forecasting is essential for power system planning and safe operation.
• Traditional grey model GM (1,1) is widely used but can be improved for accuracy.
• Existing buffer operators lack adjustable action intensity, limiting data pre-processing effectiveness.

Purpose of the Study:

• To enhance the accuracy of annual power load forecasting.
• To address the limitations of traditional weakening buffer operators in GM (1,1) models.
• To propose a novel VWWBO-BVO-based GM (1,1) model optimized for forecasting.

Main Methods:

• Dynamic pre-processing of historical power load data using Variable-Weight Weakening Buffer Operator (VWWBO) and Background Value Optimization (BVO).
• Integration of Grey Relational Analysis (GRA) and Improved Gravitational Search Algorithm (IGSA) to optimize model parameters.
• Construction of a GRA-IGSA integration algorithm to maximize grey relativity between simulated and actual values.

Main Results:

• The proposed VWWBO-BVO-based GM (1,1) model demonstrates improved forecasting accuracy.
• The GRA-IGSA optimization effectively determines optimal parameters for the buffer operator and background value.
• Case studies validate the enhanced predictive performance of the integrated model.

Conclusions:

• The VWWBO-BVO-based GM (1,1) model offers an optimized solution for annual power load forecasting.
• Adjustable buffer operator action intensity significantly improves forecasting precision.
• The GRA-IGSA integration algorithm provides robust parameter optimization for enhanced predictive accuracy.