基于K均值聚类的海上风电功率预测研究  

Research on K-means clustering based forecasting for offshore wind power

作　　者：闫健[1] YAN Jian(Personnel Office,Beijing Information Science&Technology University,Beijing 100192,China)

机构地区：[1]北京信息科技大学人事处,北京100192

出　　处：《北京信息科技大学学报（自然科学版）》2021年第4期54-58,共5页Journal of Beijing Information Science and Technology University

基　　金：促进高校内涵发展科研水平提高项目(2020KYNH211)。

摘　　要：海上风电功率预测是智能电网科学规划的先决条件,其预测误差的大小是影响电力系统调度的关键因素。通过分析海上风速、风向的变化规律与天气状况之间的关系,寻找海上风电功率的变化规律;利用K均值聚类法,构建短期风电功率预测模型的训练数据样本;利用广义回归神经网络,构建海上风电功率预测模型,该模型能够有效降低海上风电功率预测误差。最后通过具体实例验证了所构建模型的有效性。The prediction of offshore wind power forecasting is a prerequisite for the scientific planning of Smart Power Grid,and the magnitude of forecasting error is the key factor that affects the dispatch of power system.In this paper,the relationship between wind speed and wind direction and weather conditions was analyzed to find the variation of wind power at sea,and the K-means clustering method was used to build the training data sample of short-term wind power forecasting model.Based on the generalized regression neural network,a forecasting model of offshore wind power was constructed,which can reduce the forecasting error of offshore wind power effectively.Finally,an example was given to verify the validity of the model.

关 键 词：风电功率 聚类分析 短期预测 预测方法 

分 类 号：F423.2[经济管理—产业经济]