摘要
针对现有风电场功率预测方法存在预测时间短、预测精度低,不能跟踪风电波动性、间歇性进行可靠预测的问题,提出了一种基于时序指数平滑法的预测方法。该方法首先将原始数据利用指数平滑法进行去畸变量处理,得到较规则的功率数据,然后根据初步处理后的功率数据利用反馈式时序指数法进行预测。利用该方法对某大型风电场4台风电机组未来1天的发电功率进行了预测,预测结果与实测数据大致吻合,证明了该方法的可行性。
In view of problems that existing wind power prediction methods have short prediction time and low prediction accuracy,and it can＇t track characteristics of volatility and intermittent of wind power generation to get reliable prediction,the paper put forward a new prediction method based on sequential index smoothing method（SIMS）.Firstly,the method gets rid of distorted data from the original data by use of exponential smoothing method to get more ruled power data.Then it uses feedback time sequence method to predict power data.Using the method to make a prediction for the next day of four wind turbines for a large wind farm,the result is coincided with the actual measured data,which proved the method is feasibility.
出处
《工矿自动化》
北大核心
2012年第8期75-78,共4页
Journal Of Mine Automation
基金
河南省控制工程重点学科开放实验室资助项目(KG2011-15)
河南理工大学SRTP训练计划(10-3-002)
关键词
风力发电
功率预测
风速预测
时序指数平滑法
趋势效应
wind power generation
power prediction
prediction of wind speed
sequential index smoothing method
trend effect
