期刊文献+

基于聚类划分与双向LSTM网络的台区线损率计算 被引量:7

Calculation of Line Loss Rate in Transformer District Based on Cluster Division and Bidirectional LSTM Network
下载PDF
导出
摘要 在低压台区中,由于分支线路复杂,节点多,量测点少,台账数据不完整,理论线损率计算困难。提出了一种基于聚类划分与双向LSTM网络的台区线损率计算方法。首先,基于城农网标识、户均容量、运行年限等影响线损率的台区静态参数特征,采用K-medoids聚类算法将台区划分为不同类别;接着,基于台区静态参数特征以及负载率、三相不平衡度、环境温度等台区运行参数特征,采用双向LSTM网络构建每一类台区的线损率计算模型;最后,基于该模型开展台区线损率理论值计算。以某公司28167个台区样本数据进行仿真计算,结果验证了所提算法的准确性明显优于支持向量机与回归树算法的准确性。 In the low voltage transformer district,due to the complex branch lines,more nodes,not enough measurement points,incomplete equipment parameters,line loss rate calculation is difficult.A calculation method of line loss rate based on cluster division and bidirectional LSTM network is proposed.Firstly,based on the static parameters that affect the line loss rate,such as labeling of city and country power grid,average capacity per user and operating life,the K-medoids algorithm is used to divide the transformer districts into different categories.Then,based on the static parameters and operating parameters such as load rate,three-phase unbalance degree,environmental temperature and so on,the bidirectional LSTM network algorithm is used to build the calculation model of line loss rate of each type of transformer district.Finally,the theoretical value of line loss rate is calculated by the calculation model.The 28167 transformer district samples simulation results show that the accuracy of the proposed algorithm is better than that of support vector machine and regression tree algorithm.
作者 王鹏 白玉岭 王林梅 陈一鸣 高挺 孙杰 WANG Peng;BAI Yuling;WANG Linmei;CHEN Yiming;GAO Ting;SUN Jie(Taizhou Power Supply Company of State Grid,Taizhou Zhejiang 318000,China;Beijing Join Bright Digital Power Technology Co.,Ltd.,Beijing 100085,China)
出处 《电子器件》 CAS 北大核心 2022年第4期964-969,共6页 Chinese Journal of Electron Devices
基金 国网浙江省电力有限公司项目(5211DS200088)。
关键词 低压台区 线损率 静态参数 运行参数 K-medoids聚类算法 双向LSTM网络 low voltage transformer district line loss rate static parameters operating parameters K-medoids algorithm bidirectional LSTM network
  • 相关文献

参考文献15

二级参考文献127

共引文献619

同被引文献63

引证文献7

二级引证文献4

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部