Online Detection Method of CVT Error Based on Principal Component Analysis and Kernel Partial Least Squares

WANG Chunyu; YU Haibo; ZHAO Xiaojian; ZHAO Jinquan

doi:10.14044/j.1674-1757.pcrpc.2020.06.009

您当前的位置：

首页 >

文章列表页 >

Online Detection Method of CVT Error Based on Principal Component Analysis and Kernel Partial Least Squares

更新时间：2024-09-14

- Online Detection Method of CVT Error Based on Principal Component Analysis and Kernel Partial Least Squares
- POWER CAPACITOR & REACTIVE POWER COMPENSATION Vol. 41, Issue 6, Pages: 57-62(2020)
- 作者机构：
  
  1. 中国电力科学研究院
  2. 西安交通大学电气工程学院
- 作者简介：
- 基金信息：
- DOI：10.14044/j.1674-1757.pcrpc.2020.06.009
  CLC： TM451
- Published：2020，
- 稿件说明：
移动端阅览
[1]王春雨,于海波,赵晓健,赵进全.基于主元分析与核偏最小二乘的CVT误差在线检测方法[J].电力电容器与无功补偿,2020,41(06):57-62.

WANG CHUNYU, YU HAIBO, ZHAO XIAOJIAN, et al. Online Detection Method of CVT Error Based on Principal Component Analysis and Kernel Partial Least Squares. [J]. Power capacitor & reactive power compensation, 2020, 41(6): 57-62.
[1]王春雨,于海波,赵晓健,赵进全.基于主元分析与核偏最小二乘的CVT误差在线检测方法[J].电力电容器与无功补偿,2020,41(06):57-62. DOI： 10.14044/j.1674-1757.pcrpc.2020.06.009.

WANG CHUNYU, YU HAIBO, ZHAO XIAOJIAN, et al. Online Detection Method of CVT Error Based on Principal Component Analysis and Kernel Partial Least Squares. [J]. Power capacitor & reactive power compensation, 2020, 41(6): 57-62. DOI： 10.14044/j.1674-1757.pcrpc.2020.06.009.

摘要

针对CVT定期离线检测方式难于及时发现其误差变化的情况

本文提出一种基于PCAKPLS的CVT误差在线检测方法。该方法首先利用主元分析将电网波动与CVT的状态进行分离

将CVT的状态映射在残差子空间中并建立Q统计量

根据Q统计量判断CVT的状态。然后对正常运行的CVT采用核偏最小二乘法实时检测其误差。最后通过仿真

结果表明该方法能够准确地进行CVT误差检测及故障检测

是CVT误差在线检测的一种有效方法。

Abstract

In view of such situation that error variation of CVT is difficult to be found by the periodical offline detection method

a kind of online detection method of CVT error based on PCA-KPLS is proposed in this paper.The method firstly uses the principal component analysis to separate the grid fluctuation from the state of the CVT

maps the state of the CVT in the residual subspace and establishes a Q statistic and judges the state of the CVT according to the Q statistic.Then

the kernel partial least squares method in real time is used to detect the error of CVT which is in normal operation.Finally

it is shown by the simulation result that the method can perform accurately both error and fault detection of CVT and is an effective method for online detection of CVT error.

关键词

CVT误差在线检测主元分析核偏最小二乘

Keywords

CVT error online detectionprincipal component analysiskernel partial least squares

references

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

No data

Related Author

No data

Related Institution

No data

⁰