A literature review based on density forecasting and uncertainty quantification of wind power generation

Accurate density forecasting and uncertainty quantification of wind power generation are critical to the reliable integration of wind power energy into reality. These methods are used to provide the full probability distribution of wind power production while accounting for uncertainties and variability. Despite approaches to density forecasting and uncertainty quantification, forecasting wind power generation remains challenging with definitive boundaries due to inherent limitations of weather forecasts, difficulties in modeling wind, and a lack of high-quality historical data. This study reviews the literature to summarize and highlight the newest developments in wind power forecasting. Specifically, this review compiles 127 largely peer-reviewed articles published from 2010 to 2025 and analyzes available information on density forecasting for wind energy production. In this review, the methods summarized and discussed can be categorized into deterministic and probabilistic forecasting methods, focusing on short- and long-term forecasting methods. This review highlights the key advantages, disadvantages, and potential drawbacks with recommendations provided to enhance wind power generation and use of density forecasting and uncertainty quantification methods, as well as advanced data preprocessing techniques and deep learning networks (e.g., Deep Neural Network (DNN), Deep Belief Network (DBN), Convolutional Neural Network (CNN), Spiking Neural Networks (SNN)) including model configurations with the assistance of metaheuristics (e.g., Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Cuckoo Search (CS)). Furthermore, this study highlights the emerging role of Explainable Artificial Intelligence (XAI) techniques—including SHAP, LIME, and attention mechanisms—for improving model interpretability and transparency, which are vital for operational trust and decision-making in wind power systems. After thorough analysis, this paper articulates the limitations of current forecasting models and methods for wind energy production and seeks to use density forecasting frameworks and uncertainty quantification methods to improve the accuracy, reliability, and robustness of wind power forecasting systems.