风电出力短期预测技术研究-电气工程专业论文.docxVIP

华中科技 大学硕士 学 华 中 科 技 大 学 硕 士 学 位 论 文 II II Abstract Rapid global economic and social development, energy supply and demand and environmental pollution problems have become increasingly prominent, development and use of clean resources to generate sustainable development of individual countries become consensus and energy strategy. Since a huge amount of wind energy resources and clean, renewable, so the use of wind power is a kind of broad prospects for sustainable energy use. Wind power is different from the conventional power of randomness and other characteristics, wind power capacity increased significantly affect power system security, stability and economic operation. Achieve higher accuracy in favor of wind power output forecast grid scheduling and network management, and to ensure the safe operation of the system, improve the system capacity of wind power and wind power consumptive bidding capabilities. Firstly, the development of wind power industry at home and abroad as well as wind power output prediction technology research status summary and conclusion. Taking into account the statistical method is wind power output and short-term forecasting of the most widely used prediction methods, and time series forecasting model and support vector machine model is a statistical method more typical prediction model, this paper based on time series method and support vector machine theory respectively. established based on the time series method based on support vector machine and wind power output short-term forecasting model, using the built forecasting models are examples and comparative analysis of the characteristics of two prediction models. Traditional statistical forecasting methods by building between input and output parameters of the function to make predictions work due to the randomness of wind power output and complexity, the establishment of parametric predictive model is not necessarily the best choice. Through further analysis of wind power output