考虑风电场功率爬坡的超短期组合预测  

An Ultra-Short Term Combined Prediction Considering Wind Farm Power Climbing

作　　者：杨茂[1] 于欣楠 YANG Mao;YU Xinnan(Key Laboratory of Modern Power System Simulation and Control&Renewable Energy Technology,Ministry of Education Northeast Electric Power University,Jilin 132012)

机构地区：[1]东北电力大学现代电力系统仿真控制与绿色电能新技术教育部重点实验室,吉林132012

出　　处：《东北电力大学学报》2022年第1期63-70,共8页Journal of Northeast Electric Power University

基　　金：国家重点研发计划项目促进可再生能源清纳的风电光伏发电功率预测技术及应用(2018YFB0904200)资助。

摘　　要：随着大规模风力发电接入电力系统,准确的风电场功率预测对于整个电力系统的安全稳定运行均意义重大.而风电功率爬坡事件则是产生风电功率预测误差的重要原因,尤其是当风速数据变化较快时,所引发的功率爬坡会导致预测误差较大.因此研究考虑风电场功率爬坡事件的预测就显得日益紧迫.文中基于极限学习机理论,提出了一种考虑风电场功率爬坡的超短期组合预测模型.经算例验证表明,文中方法能够准确识别风电场的功率爬坡事件并有效提高风电功率超短期预测的精度,具有一定的理论意义和实用价值.With the large-scale application of wind power in the power system,the precise prediction of wind power is of great significance for the stable and safe operation of the power system.The wind power climbing event is an important reason for the prediction error of wind power.Especially when wind speed changes rapidly,the prediction error caused by power climbing event is considerably large.Therefore,it is increasingly urgent to study how to predict the wind power considering the wind farm power climbing event.Based on the theory of extreme learning machine,a super short-term power combination prediction model considering wind power climbing event is proposed in this paper.The results show that the proposed method can accurately identify wind power climbing events and effectively improve the prediction accuracy of ultra-short-term wind power prediction,which has certain theoretical significance and practical value.

关 键 词：风电爬坡事件 极限学习机 数值天气预报 风电功率预测 

分 类 号：TM614[电气工程—电力系统及自动化]