Research On The Wind Power Characteristics Of Wind Farms Based On Measured Data

With the fossil fuels depletion, global climate change and serious environmental pollution, the utilization of renewable energy has become the general trend of world energy development. Wind energy is one of the most promising new energy. Its stochastic characteristics and volatility are great challenges to the power system operation. Based on the wind power measured data, this thesis has done a lot of research on wind power fluctuation in different time and space scale, wind power correlation and characteristics of predicted wind power in detail in order to support the power system dispatching and reduce wind curtailment. The main work is as follows:1) Wind power bad data identification based on outlier detection has been proposed. The algorithm has been validated using simulation data and measured wind power data of Inner Mongolia.2) Based on measured data of Inner Mongolia wind power base, the probability distribution of wind power fluctuation and peak-valley value in different time and space scales has been investigated. The correlation coefficient with and without considering time delay has been calculated. The algorithm of correlation of average wind speed based on wind power has been given and applied in Jilin wind power base.3) The method of characteristics of predicted wind power based on three-parameter Weibull distribution has been presented. Quantitative analysis and comparison of the three-parameter Weibull distribution and the general distribution, normal distribution and Beta distribution have been realized with field data of Inner Mongolia wind power base.