摘要
为了充分挖掘和高效利用新能源的消纳空间,需要在调度控制时考虑广域量测、环境、气象以及社会等多源信息。数据清洗是使用这些数据的前提,提出了一种数据清洗技术,分析了新能源消纳能力计算的数据来源以及数据间的隐含关联,将其转换为PSL形式的模糊规则,构建了面向新能源消纳数据清洗的神经网络。以张家口市张北县2020年风电数据为例进行数据清洗。结果表明:数据清洗并修复后的数据正确率达到98.3%,该技术能够有效地利用多源信息的关联映射关系,实现对不良数据的有效甄别。研究结果可为各方参与新能源建设和开展新能源价值挖掘提供可靠的数据支撑。
To fully exploit and efficiently utilize the integration capacity of new energy,it is essential to consider a range of information sources,including extensive measurements,environmental factors,weather conditions,and societal aspects in the scheduling and control processes and data cleaning serves as a prerequisite for utilizing this information.Therefore,we proposed a data cleaning technique that analyzed the data sources and hidden correlations in the calculation of new energy integration capacity,transforming them into fuzzy rules represented in the form of Probabilistic Soft Logic.We constructed a neural network,specifically designed for data cleaning in new energy integration.We applied it to the wind power data from Zhangbei County,Zhangjiakou City in 2020,achieving a data cleaning accuracy rate of 98.3%.The results showed that the technique was able to effectively utilize the associative mapping relationship among multiple sources of information,thereby enabled the identification of faulty data.The research results can provide a data support for various stakeholders involved in the development and exploration of new energy solutions.
作者
熊为军
肖华
陈卫鹏
XIONG Weijun;XIAO Hua;CHEN Weipeng(Changjiang Survey,Planning,Design and Research Co.,Ltd.,Wuhan 430010,China)
出处
《水利水电快报》
2024年第7期116-121,共6页
Express Water Resources & Hydropower Information
关键词
新能源消纳
数据清洗
模糊规则
概率软逻辑
new energy consumption
data cleaning
fuzzy rules
probabilistic soft logic
作者简介
熊为军,男,高级工程师,主要从事水电站电气一次设计、电网规划和输变电设计工作。E-mail:xiongweijun@cjwsjy.com.cn。
