Wind Power Cluster Power Prediction Based On Cluster Analysis

Nowadays,the rapid development of the economy mad the energy consumption rate continue to accelerate.In order to prevent human society from the dilemma of energy exhaustion,countries around the world had been successively promoting the research and application of clean and sustainable energy.Wind energy is a renewable energy with a relatively mature level of research and development,and the installed capacity of wind power in the world is constantly increasing.Although the technology of wind power is increasingly developed,the development of wind power is restricted due to the uneven distribution of wind resources,the uneven development of the industrial chain,and the insufficient single unit capacity of wind turbines.In particular,wind power often exhibits characteristics as intermittent and random.While the installed capacity of wind power is increasing,the expansion of the scale of wind power grid integration has also adversely affected the safe and stable operation of power systems and the guarantee of power generation quality.In order to use the clean energy as wind more effectively,it is necessary to make effective predictions of wind power.This can help the operation di spatcher to formulate a reasonable operation plan and adjust it in a timely manner,thereby increasing the system’s wind power acceptance rate and reducing the situation of "abandoning wind and limiting power".This article elaborated the current development status of wind power in the world and China,put forward the advantages and necessity of developing wind power,and analyzed the current research status of power prediction technologies for wind farms and wind power clusters.This paper learned and anal yzed the output characteristics of wind power,and proposed corresponding indicators for its time and space characteristics.Based on the theoretical knowledge,the power prediction method of wind power clusters based on cluster analysis was introduced.The clustering algorithm was used to divide the wind power cluster into several sub-regions,and the representative wind farm was selected according to the correlation and prediction accuracy of the wind farm.Then,the prediction results of the representative wind farm were weighted and summed using the upscaling algorithm,and finally the power prediction value of the cluster.Using this method to carry out simulation experiments on historical data of wind farms,the results can predict the cluster power more accurately,indicating that the prediction results of this method are more accurate.