Adaptive Kernel Density State Estimation Considering Leverage and Strongly Influential Bad Data Identification

ZHU Kaiyu; LIN Jikeng

doi:10.7500/AEPS20250214001

您当前的位置：

首页 >

文章列表页 >

Adaptive Kernel Density State Estimation Considering Leverage and Strongly Influential Bad Data Identification

Basic Research | 更新时间：2026-02-28

- Adaptive Kernel Density State Estimation Considering Leverage and Strongly Influential Bad Data Identification
- Automation of Electric Power Systems Vol. 49, Issue 19, Pages: 164-173(2025)
- 作者机构：
  
  同济大学电子与信息工程学院，上海市 201804
- 作者简介：
- 基金信息：
  
  National Natural Science Foundation of China(51177107)
- DOI：10.7500/AEPS20250214001
  CLC：
- Received：14 February 2025，
  
  Revised：2025-06-10，
  
  Online First：29 July 2025，
  
  Published：10 October 2025
- 稿件说明：
移动端阅览
朱锴宇,林济铿.考虑杠杆及强影响不良数据辨识的自适应核密度状态估计[J].电力系统自动化,2025,49(19):164-173.

ZHU Kaiyu,LIN Jikeng.Adaptive Kernel Density State Estimation Considering Leverage and Strongly Influential Bad Data Identification[J].Automation of Electric Power Systems,2025,49(19):164-173.
朱锴宇,林济铿.考虑杠杆及强影响不良数据辨识的自适应核密度状态估计[J].电力系统自动化,2025,49(19):164-173. DOI： 10.7500/AEPS20250214001.

ZHU Kaiyu,LIN Jikeng.Adaptive Kernel Density State Estimation Considering Leverage and Strongly Influential Bad Data Identification[J].Automation of Electric Power Systems,2025,49(19):164-173. DOI： 10.7500/AEPS20250214001.

摘要

电力系统量测数据中存在杠杆及强影响不良数据，因其难以辨识，导致状态估计结果存在偏差。为提高状态估计精度，提出改进的自适应核密度状态估计方法。该方法采用投影统计量与标准化残差，实现杠杆及非杠杆不良数据的准确辨识；然后，基于不良数据、杠杆及强影响数据的辨识结果，通过自适应调整量测权重，并优化量测核带宽的计算，剔除不良数据影响。算例分析结果表明，所提方法可以显著降低错误辨识率，且其计算速度可满足工程实时性要求。

Abstract

The power system measurements have leverage and strongly influential bad data

which causes deviations in state estimation results because they are difficult to identify. To improve the accuracy of state estimation

an improved adaptive kernel density state estimation method is proposed. This method employs the projection statistics and standardized residuals to achieve the accurate identification of both leverage and non-leverage bad data. Subsequently

based on the identification results of bad data

leverage and strongly influential data

the measurement weights are adaptively adjusted

and the calculation of measurement kernel bandwidth is optimized to eliminate the influences of bad data. Case study results demonstrate that the proposed method can significantly reduce misidentification rates

and its calculation speed can meet the real-time engineering requirements.

关键词

Keywords

references

ZHAO J B ， MILI L . Vulnerability of the largest normalized residual statistical test to leverage points ［J］. IEEE Transactions on Power Systems ， 2018 ， 33 （ 4 ）： 4643 - 4646 .

WEI S H ， XU J J ， WU Z J ， et al . A false data injection attack detection strategy for unbalanced distribution networks state estimation ［J］. IEEE Transactions on Smart Grid ， 2023 ， 14 （ 5 ）： 3992 - 4006 .

ABUR A ， EXPOSITO A G . Power system state estimation： theory and implementation ［M］. New York， USA ： Marcel Dekker Incorporation ， 2004 .

ZHANG H Q ， XU Q ， XIE Y ， et al . Robust state estimation method based on Mahalanobis distance under non-Gauss noise ［J］. IEEE Access ， 2024 ， 12 ： 9243 - 9250 .

ABOLMASOUMI A H ， FARAHANI A ， MILI L . Robust particle filter design with an application to power system state estimation ［J］. IEEE Transactions on Power Systems ， 2024 ， 39 （ 1 ）： 1810 - 1821 .

刘有志，资慧，肖健，等 . 杠杆量测内涵分析及其改进广义潜能指标辨识方法［J］. 电力系统自动化， 2021 ， 45 （ 12 ）： 92 - 100 .

LIU Youzhi ， ZI Hui ， XIAO Jian ， et al . Connotation analysis of leverage measurement and its identification method based on improved generalized potential indicator ［J］. Automation of Electric Power Systems ， 2021 ， 45 （ 12 ）： 92 - 100 .

RAHMATULLAH IMON A H M . Identifying multiple influential observations in linear regression ［J］. Journal of Applied Statistics ， 2005 ， 32 （ 9 ）： 929 - 946 .

CHEN Y B ， LIU F ， MEI S W ， et al . A robust WLAV state estimation using optimal transformations ［J］. IEEE Transactions on Power Systems ， 2015 ， 30 （ 4 ）： 2190 - 2191 .

DOĞRU F Z ， ARSLAN O . Robust mixture regression modeling based on the generalized M （GM）-estimation method ［J］. Communications in Statistics—Simulation and Computation ， 2021 ， 50 （ 9 ）： 2643 - 2665 .

DOBAKHSHARI A S ， ABDOLMALEKI M ， TERZIJA V ， et al . Robust hybrid linear state estimator utilizing SCADA and PMU measurements ［J］. IEEE Transactions on Power Systems ， 2021 ， 36 （ 2 ）： 1264 - 1273 .

RESENDE E M ， PIRES R ， MILI L . Complex-valued projection statistics for hybrid power system state estimation ［J］. IEEE Transactions on Power Systems ， 2024 ， 39 （ 4 ）： 6052 - 6062 .

MAJUMDAR A ， PAL B C . Bad data detection in the context of leverage point attacks in modern power networks ［J］. IEEE Transactions on Smart Grid ， 2018 ， 9 （ 3 ）： 2042 - 2054 .

LIN J K ， YANG C Q ， LIU Y S ， et al . Practical state estimation model and algorithm based on kernel density theory ［J］. International Journal of Electrical Power & Energy Systems ， 2022 ， 134 ： 107442 .

BRETAS N G ， BRETAS A S . A two steps procedure in state estimation gross error detection， identification， and correction ［J］. International Journal of Electrical Power & Energy Systems ， 2015 ， 73 ： 484 - 490 .

安装式数字显示电测量仪表第3部分：功率表和无功功率表的特殊要求： GB/T 22264.3—2022 ［S］. 北京：中国标准出版社， 2022 .

Mounted digital display electric measuring instruments： Part 3： special requirements for wattmeters and varmeters ： GB/T 22264.3—2022 ［S］. Beijing ： Standards Press of China ， 2022 .

WANG X Z ， LIU Y T ， LI Z S ， et al . Enhanced maximum exponential square state estimator based on simulation-hyperparameter optimization ［J］. IEEE Transactions on Power Systems ， 2023 ， 38 （ 6 ）： 5957 - 5960 .

CHEN Y B ， MA J ， ZHANG P ， et al . Robust state estimator based on maximum exponential absolute value ［J］. IEEE Transactions on Smart Grid ， 2017 ， 8 （ 4 ）： 1537 - 1544 .

钱峰，龚庆武 . 基于IGG法的电力系统状态估计［J］. 电力系统自动化， 2005 ， 29 （ 3 ）： 36 - 39 .

QIAN Feng ， GONG Qingwu . State estimation of power system based on IGG method ［J］. Automation of Electric Power Systems ， 2005 ， 29 （ 3 ）： 36 - 39 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Dynamic State Estimation Method for Harmonics in AC/DC Distribution Network Based on KalmanNet

Joint Estimation of Lithium-ion Battery State Based on Light Gradient Boosting Machine

Distribution Network Restoration Decision Considering Risk of State Estimation Error in Restoration Process

Related Author

ZHU Kaiyu

LIN Jikeng

LIU Jinbo

GUO Qinglai

ZHANG Jing

YAN Yaqin

WANG Miao

SUN Lue

Related Institution

College of Electronics and Information Engineering, Tongji University

National Power Dispatching and Control Center, State Grid Corporation of China

Department of Electrical Engineering, Tsinghua University

National Key Laboratory of Grid Safety (China Electric Power Research Institute)

School of Electrical and Power Engineering, Hohai University

⁰