A Survey of Computational Intelligence Techniques for Wind Power Uncertainty Quantification in Smart Grids

Area of Science:

• Electrical Engineering
• Computer Science
• Energy Systems

Background:

• High penetration of renewable energy sources like wind power introduces significant uncertainties into power systems.
• Traditional power system decision-making methods are inadequate for managing these wind power uncertainties.
• Smart grids require advanced techniques to handle the variability and unpredictability of wind energy.

Purpose of the Study:

• To provide a comprehensive survey of computational intelligence techniques for wind power uncertainty quantification in smart grids.
• To review methods for incorporating these uncertainties into power system decision-making processes.
• To identify future research directions in this critical area.

Main Methods:

• Quantifying wind power forecast uncertainties using prediction intervals (PIs) and evaluating various PI construction methods (parametric and nonparametric).
• Investigating techniques for integrating wind power uncertainties into decision-making, including stochastic models, fuzzy logic models, and robust optimization.
• Reviewing diverse power system applications that utilize these uncertainty management techniques.

Main Results:

• Comparison of different PI evaluation indices and PI construction methods for wind power forecasting.
• Overview of how stochastic models, fuzzy logic, and robust optimization address wind power uncertainties in power system operations.
• Identification of emerging trends and challenges in the field.

Conclusions:

• Computational intelligence offers powerful tools for managing wind power uncertainties in smart grids.
• Effective uncertainty quantification and integration into decision-making are crucial for reliable smart grid operation.
• Future research should focus on advanced forecasting, deep learning, and seamless integration of uncertainty estimates.