摘要
在研究风电场或风电机组出力特性前,对数据的预处理很必要。数据可能出现的问题较多,集中的大量数据缺失是其中比较棘手的问题之一。针对这个问题,在分析风电场或风电机组之间时间和空间的分布特性的基础上,建立对集中的大量数据缺失补齐的时空模型,并根据时间模型和空间模型的特点对建立的时空模型进行改善,提出了一种双向变权重的时空模型。以均方根误差为指标,评价双向变权重的时空模型与单一时间、空间模型的补齐精度。通过算例验证双向变权重的时空模型优于单一模型,补齐精度有明显的提高。
The data preprocessing is necessary before res- earches on the output characteristics of wind farms or wind turbines are conducted. Of the problems that may occur, lack of adequate data is a very crucial one. This paper analyzes the distribution space-time characteristics of the wind farms or wind turbines and then establishes the space-time model for polishing the concentrated and mass missing data. Based on the features of the time and space, the paper improves the space- time model and comes up with the bidirectional variable weight space-time model. Taking the root-mean-square error as the indicator, it evaluates the polishing precisions of the bidirectional variable weight space-time model, the single time and the single space model. Finally, through the simulation example, the conclusion is verified that the bidirectional variable weight space-time model is superior to the single time or space model and improves the polishing precision.
出处
《电网与清洁能源》
2013年第9期74-80,共7页
Power System and Clean Energy
基金
国家高技术研究发展计划(863计划)(2012AA050203)~~
关键词
数据预处理
回归模型
风电功率时空特性
双向变权重时空模型
data pre-processing
regression model
wind power in space-time characteristics
bidirectional variable weight space-time model
