分布式发电系统中机器学习实时应用
发布时间:2018-09-13 12:43
【摘要】:提出聚类采样预处理和集成分类器的方法来应对含概念漂移的分布式发电孤岛检测问题。首先提出利用聚类算法对电气信息进行预处理和再采样。然后分析了单一分类器的归纳偏置现象,提出利用集成分类器的互补性提高孤岛检测的准确率。仿真算例中充分考虑了功率不平衡度,本地负荷扰动等因素。实验结果表明,上述两个环节对提高分布式发电孤岛检测的精度和泛化能力具有重要作用。
[Abstract]:A method of cluster sampling preprocessing and integrated classifier is proposed to deal with the problem of distributed generation islanding detection with concept drift. Firstly, a clustering algorithm is proposed to preprocess and resample electrical information. Then the inductive bias phenomenon of single classifier is analyzed and the complementarity of integrated classifier is proposed to improve the accuracy of islanding detection. The power imbalance and local load disturbance are fully considered in the simulation examples. The experimental results show that the above two links play an important role in improving the accuracy and generalization ability of distributed islanding detection.
【作者单位】: 上海电力设计院有限公司;国网上海市电力公司;
【分类号】:TM61
[Abstract]:A method of cluster sampling preprocessing and integrated classifier is proposed to deal with the problem of distributed generation islanding detection with concept drift. Firstly, a clustering algorithm is proposed to preprocess and resample electrical information. Then the inductive bias phenomenon of single classifier is analyzed and the complementarity of integrated classifier is proposed to improve the accuracy of islanding detection. The power imbalance and local load disturbance are fully considered in the simulation examples. The experimental results show that the above two links play an important role in improving the accuracy and generalization ability of distributed islanding detection.
【作者单位】: 上海电力设计院有限公司;国网上海市电力公司;
【分类号】:TM61
【参考文献】
相关期刊论文 前2条
1 林霞;董晓峰;陆于平;刘玉欢;;基于数据挖掘的分布式系统孤岛检测方法[J];电力系统及其自动化学报;2011年02期
2 王涛;李舟军;颜跃进;陈火旺;;数据流挖掘分类技术综述[J];计算机研究与发展;2007年11期
【共引文献】
相关期刊论文 前10条
1 李广水;宋丁全;;数据分析在森林资源调查中的应用及发展研究[J];安徽农业科学;2009年22期
2 崔荣靖;李盛伟;李晓辉;王哲;;光伏并网逆变器检测技术初探[J];电力系统及其自动化学报;2013年04期
3 肖龙;杨国华;周鑫;姚琪;朱向芬;智腾飞;王金梅;;光伏微网系统孤岛检测中的阻抗测量仿真[J];电力系统及其自动化学报;2013年05期
4 马保雷;宋颖慧;刘亚维;;基于概念漂移检测的自适应流量识别的研究[J];智能计算机与应用;2013年06期
5 张沛超;谭啸风;杨s铞,
本文编号:2241205
本文链接:https://www.wllwen.com/kejilunwen/dianlilw/2241205.html