曹渝昆,胡清清.计及故障时间的神经网络风力发电量预测[J].电气自动化,2017,(5):45~47, 55
计及故障时间的神经网络风力发电量预测
Neural Network Forecast of Wind Power Generation under Consideration of Fault Time
  修订日期:2016-12-10
DOI:
中文关键词:  风电发电量预测  时间序列  神经网络  故障时间  气象因素
英文关键词:wind power generation forecast  time series  neural network  fault time  meteorological factor
基金项目:
作者单位
曹渝昆 上海电力学院计算机与科学技术学院, 上海 200090 
胡清清 上海电力学院计算机与科学技术学院, 上海 200090 
摘要点击次数: 936
全文下载次数: 
中文摘要:
      风电是可再生能源的一种重要形式,随着越来越多的风电并入电网,发电量的预测对电网的稳定性变得格外重要,为了得到更高的预测精度,提出了一种将天气因素和故障时间相结合的两段式风力发电量预测的方法。结合湖南省某风力发电场的实际生产数据,分别运用神经网络结合时间序列的方法对故障时间进行预测,以及GRNN神经网络方法对发电量进行预测。提高了风力发电量预测的精度,延长了预测时间,证明了方法的可行性。
英文摘要:
      Wind power is an important form of renewable energy. While more and more wind power is integrated into the grid, power generation forecasting becomes extraordinarily important to grid stability. To achieve a higher prediction accuracy, this paper proposes a two-section method for wind power generation forecast which combines weather factors with fault time. Under consideration of actual production data from a wind farm in Hunan Province, fault time is predicted in the combined way of neural network and time series while generating capacity is predicted in the GRNN neural network method, so that forecast accuracy of wind power generation is improved, forecast time is elongated, and the feasibility of this approach is verified.
查看全文  查看/发表评论  下载PDF阅读器
关闭