An actual load forecasting methodology by interval grey modeling based on the fractional calculus

Area of Science:

• Engineering
• Data Science
• Applied Mathematics

Background:

• Thermal power plant operations generate extensive real-time and historical interval data crucial for decision-making and maintenance.
• Actual load is a key parameter reflecting equipment status, but modeling interval grey numbers presents challenges.
• Existing methods for interval data modeling can be complex and may lose information.

Purpose of the Study:

• To develop a novel grey prediction model for interval grey numbers using fractional-order accumulation calculus.
• To address the complexities of modeling and predicting with interval data in thermal power plant operations.
• To preserve all information from original interval data by utilizing geometric coordinate features.

Main Methods:

• Utilizing geometric coordinate features (area and middle point lines) to represent interval data without information loss.
• Proposing a fractional-order grey prediction model based on accumulation calculus for interval grey numbers.
• Comparing the performance of the fractional-order model against integer-order models.

Main Results:

• Geometric coordinate features effectively retain the information of original interval data.
• The proposed fractional-order grey prediction model demonstrates improved performance in modeling and prediction.
• The method shows greater flexibility compared to traditional integer-order models.

Conclusions:

• The fractional-order grey prediction model offers a robust approach for analyzing interval data in thermal power plants.
• This method is suitable for modeling and prediction with limited historical industrial sequence samples.
• The technique enhances the understanding of equipment operation status and supports predictive maintenance strategies.