A Review on the Temporal-Spatial Correlation Based Multiple Horizon Wind Power Forecasting Methods
摘要
Among various renewable energy sources such as solar photovoltaic, tidal and waves, wind power (WP) has taken the lead due to its high efficiency, affordability, and abundance. However, the power generated by wind turbines can vary rapidly due to fluctuations in wind speed and direction. It is also influenced by terrain, humidity, time of day, and date. The random nature of wind speed poses challenges to the safety and stability of electric power grids when WP is integrated into grids on large scales. Therefore, WP forecasting is a method to decrease the risk of uncertainties in wind, allowing higher penetration. This paper reviews various temporal-spatial methods, compares different forecasting methods, and summarizes future research directions as a guide to improve the accuracy of wind power forecasting (WPF).