Bayesian state estimation for partially observable distribution networks via power flow-informed neural networks

The performance of existing distribution network state estimation (SE) methods is unsatisfactory due to limited real-time measurements. In this paper, a Bayesian SE method is proposed for partially observable distribution networks using a novel power flow-informed neural network (PFINN). The Bayesia...

Full description

Saved in:
Bibliographic Details
Main Authors: Dong Liang, Guirong Li, Xiaoyu Liu, Lin Zeng, Hsiao-Dong Chiang, Shouxiang Wang
Format: Article
Language:English
Published: Elsevier 2025-09-01
Series:International Journal of Electrical Power & Energy Systems
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S014206152500434X
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849390111052005376
author Dong Liang
Guirong Li
Xiaoyu Liu
Lin Zeng
Hsiao-Dong Chiang
Shouxiang Wang
author_facet Dong Liang
Guirong Li
Xiaoyu Liu
Lin Zeng
Hsiao-Dong Chiang
Shouxiang Wang
author_sort Dong Liang
collection DOAJ
description The performance of existing distribution network state estimation (SE) methods is unsatisfactory due to limited real-time measurements. In this paper, a Bayesian SE method is proposed for partially observable distribution networks using a novel power flow-informed neural network (PFINN). The Bayesian SE model is first established with the goal of directly minimizing the SE error and obtaining the expectation of states conditional on measurements. The conditional expectation is then viewed as a nonparametric regression that can be parameterized by a PFINN where a physics loss penalty is introduced into the loss function to constrain the neural network outputs to be consistent with power network operating constraints. Test results on balanced and unbalanced distribution networks using field data show that the proposed method can achieve better estimation accuracy than the Bayesian SE method without power flow informing. For a fixed neural network, it is always possible to enhance its performance by adding a physics loss term to the original loss function, along with a suitable weight which can be obtained using existing automatic hyperparameter tunning methods.
format Article
id doaj-art-b072bd068db7452289ba062fd7b591fc
institution Kabale University
issn 0142-0615
language English
publishDate 2025-09-01
publisher Elsevier
record_format Article
series International Journal of Electrical Power & Energy Systems
spelling doaj-art-b072bd068db7452289ba062fd7b591fc2025-08-20T03:41:46ZengElsevierInternational Journal of Electrical Power & Energy Systems0142-06152025-09-0117011088610.1016/j.ijepes.2025.110886Bayesian state estimation for partially observable distribution networks via power flow-informed neural networksDong Liang0Guirong Li1Xiaoyu Liu2Lin Zeng3Hsiao-Dong Chiang4Shouxiang Wang5Innovation Research Institute of Hebei University of Technology, Shijiazhuang 050222, China; State Key Laboratory of Intelligent Power Distribution Equipment and System, Hebei Key Laboratory of Equipment and Technology Demonstration of Flexible DC Transmission, Department of Electrical Engineering, Hebei University of Technology, Tianjin 300401, ChinaInnovation Research Institute of Hebei University of Technology, Shijiazhuang 050222, China; State Key Laboratory of Intelligent Power Distribution Equipment and System, Hebei Key Laboratory of Equipment and Technology Demonstration of Flexible DC Transmission, Department of Electrical Engineering, Hebei University of Technology, Tianjin 300401, ChinaState Grid Tangshan Power Supply Company, Tangshan 063000, ChinaSchool of Electrical and Computer Engineering, Cornell University, Ithaca, NY 14853, USA; Corresponding author.School of Electrical and Computer Engineering, Cornell University, Ithaca, NY 14853, USAKey Laboratory of Smart Grid of Ministry of Education, Tianjin University, Tianjin 300072, ChinaThe performance of existing distribution network state estimation (SE) methods is unsatisfactory due to limited real-time measurements. In this paper, a Bayesian SE method is proposed for partially observable distribution networks using a novel power flow-informed neural network (PFINN). The Bayesian SE model is first established with the goal of directly minimizing the SE error and obtaining the expectation of states conditional on measurements. The conditional expectation is then viewed as a nonparametric regression that can be parameterized by a PFINN where a physics loss penalty is introduced into the loss function to constrain the neural network outputs to be consistent with power network operating constraints. Test results on balanced and unbalanced distribution networks using field data show that the proposed method can achieve better estimation accuracy than the Bayesian SE method without power flow informing. For a fixed neural network, it is always possible to enhance its performance by adding a physics loss term to the original loss function, along with a suitable weight which can be obtained using existing automatic hyperparameter tunning methods.http://www.sciencedirect.com/science/article/pii/S014206152500434XPower flow-informed neural networkBayesian state estimationDistribution network
spellingShingle Dong Liang
Guirong Li
Xiaoyu Liu
Lin Zeng
Hsiao-Dong Chiang
Shouxiang Wang
Bayesian state estimation for partially observable distribution networks via power flow-informed neural networks
International Journal of Electrical Power & Energy Systems
Power flow-informed neural network
Bayesian state estimation
Distribution network
title Bayesian state estimation for partially observable distribution networks via power flow-informed neural networks
title_full Bayesian state estimation for partially observable distribution networks via power flow-informed neural networks
title_fullStr Bayesian state estimation for partially observable distribution networks via power flow-informed neural networks
title_full_unstemmed Bayesian state estimation for partially observable distribution networks via power flow-informed neural networks
title_short Bayesian state estimation for partially observable distribution networks via power flow-informed neural networks
title_sort bayesian state estimation for partially observable distribution networks via power flow informed neural networks
topic Power flow-informed neural network
Bayesian state estimation
Distribution network
url http://www.sciencedirect.com/science/article/pii/S014206152500434X
work_keys_str_mv AT dongliang bayesianstateestimationforpartiallyobservabledistributionnetworksviapowerflowinformedneuralnetworks
AT guirongli bayesianstateestimationforpartiallyobservabledistributionnetworksviapowerflowinformedneuralnetworks
AT xiaoyuliu bayesianstateestimationforpartiallyobservabledistributionnetworksviapowerflowinformedneuralnetworks
AT linzeng bayesianstateestimationforpartiallyobservabledistributionnetworksviapowerflowinformedneuralnetworks
AT hsiaodongchiang bayesianstateestimationforpartiallyobservabledistributionnetworksviapowerflowinformedneuralnetworks
AT shouxiangwang bayesianstateestimationforpartiallyobservabledistributionnetworksviapowerflowinformedneuralnetworks