Abstract：With the rapid advancement of wind and solar power generation technologies, the proportion of wind and solar energy within power systems continues to rise. However, the output of these two energy sources is directly influenced by weather conditions, exhibiting significant randomness, volatility, and intermittency. This poses severe challenges to power dispatch and the secure, stable operation of the grid. High-precision power forecasting technologies at short- and medium-term time scales are crucial for addressing these challenges and achieving efficient utilization of wind and solar power generation. This paper systematically reviews the developmental trajectory, core technologies, and latest research advancements in short- and medium-term meteorological forecasting for wind and solar energy. First, a comprehensive analysis of relevant domestic and international literature indicates that traditional wind and solar energy forecasting methods—such as numerical weather prediction techniques, statistical approaches (including time series analysis and machine learning), and hybrid/statistical post-processing methods combining both—no longer fully meet current demands. Significant improvements in forecast accuracy and reliability have been achieved through synergistic optimization in key areas: high-resolution rapid-cycle forecasting, energy-specific physical process optimization, and hybrid data assimilation method of ensemble Kalman filtering (EnKF) and four-dimensional variational assimilation (4DVar). Furthermore, the introduction of next-generation dynamic frameworks and advanced artificial intelligence (AI) large models, the assimilation and fusion of multi-source data, and the application of emerging technologies like “digital twins” offer new avenues for further refining forecast outcomes. Finally, the paper analyses the challenges confronting current forecasting techniques and outlines future development directions which include but not limited to challenges posed by forecasting extreme weather events， complex terrain physical and dynamical representations, enhancing model interpretability, and subseasonal to seasonal-scale forecasting; Deep integration of artificial intelligence with physical models to develop “physical-information neural networks” by embedding physical laws; Collective probabilistic forecasting that combines traditional physical models with large AI model ensembles will become widespread; Disruptive technologies like quantum computing are poised to advance ultra-high-resolution meteorological simulations; Achieving deep coupling between forecasting technologies and scenarios such as the grid-based automatic power generation control to build closed-loop intelligent decision-making systems. This review aims to provide a technical reference for researchers and engineering technicians in the field of energy meteorology.