Physics-Informed Machine Learning for Data Anomaly Detection, Classification, Localization, and Mitigation - A Review, Challenges, and Path Forward

Aug 5, 2026

Upload by Finn

Advancements in digital automation for smart grids deploy measurement devices such as PMUs, micro-PMUs, and smart meters, generating large-scale data that control-room operators must translate into reliable, resilient cyber-power decisions. Physics-informed machine learning (PIML) supports efficient, trustworthy, and interpretable models by embedding power-grid physical principles to address limitations of purely data-driven approaches. The work reviews strategies across supervised, semi-supervised, unsupervised, and reinforcement learning, surveys PIML for anomaly detection, classification, localization, and mitigation, and discusses improvements and real-world suitability.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages19

Reading Time48 min

AccessFree to download

Abstract

Index Terms

I. Introduction

Physics-Informed Machine Learning for Data Anomaly Detection, Classification, Localization, and Mitigation: A Review, Challenges, and Path Forward Mehdi Jabbari Zideh, Student Member, IEEE, Paroma Chatterjee, Member, IEEE, and Anurag K. Srivastava, Fellow, IEEE arXiv :2309 . 10788v1 [ ee ss . SY] 19 Sep 2023 Abstract—dvancements in digital automation for smart grids have led to the installation of measurement devices like phasor measurement units (PMUs), micro-PMUs (µ-PMUs), and smart meters. However, a large amount of data collected by these devices brings several challenges as control room operators need to use this data with models to make confident decisions for reliable and resilient operation of the cyber-power systems. Machinelearning (ML) based tools can provide a reliable interpretation of the deluge of data obtained from the field. For the decisionmakers to ensure reliable network operation under all operating conditions, these tools need to identify solutions that are feasible and satisfy the system constraints, while being efficient, trustworthy, and interpretable. This resulted in the increasing popularity of physics-informed machine learning (PIML) approaches, as these methods overcome challenges that model-based or datadriven ML methods face in silos. This work aims at the following: a) review existing strategies and techniques for incorporating underlying physical principles of the power grid into different types of ML approaches (supervised/semi-supervised learning, unsupervised learning, and reinforcement learning (RL)); b) explore the existing works on PIML methods for anomaly detection, classification, localization, and mitigation in power transmission and distribution systems, c) discuss improvements in existing methods through consideration of potential challenges while also addressing the limitations to make them suitable for real-world applications.dvancements in digital automation for smart grids have led to the installation of measurement devices like phasor measurement units (PMUs), micro-PMUs (µ -PMUs), and smart meters. However, a large amount of data collected by these devices brings several challenges as control room operators need to use this data with models to make confident decisions for reliable and resilient operation of the cyber-power systems. Machine-learning (ML) based tools can provide a reliable interpretation of the deluge of data obtained from the field. For the decision-makers to ensure reliable network operation under all operating conditions, these tools need to identify solutions that are feasible and satisfy the system constraints, while being efficient, trustworthy, and interpretable. This resulted in the increasing popularity of physics-informed machine learning (PIML) approaches, as these methods overcome challenges that model-based or data-driven ML methods face in silos. This work aims at the following: a) review existing strategies Authors are with the Lane Department of Computer Science and Electrical Engineering, West Virginia University, Morgantown, WV 26505 USA (e-mail: [mj00021@mix.wvu.edu](mj00021@mix.wvu.edu), [pc00026@mix.wvu.edu](pc00026@mix.wvu.edu), [anurag.srivastava@mail.wvu.edu](anurag.srivastava@mail.wvu.edu)) This work was supported in part by the U.S. National Science Foundation FW-HTF award 1840192 . We would like to acknowledge Dr. Sarika Khushalani Solanki for technical support. Corresponding author: Mehdi Jabbari Zideh (e-mail: [mj00021@mix.wvu.edu](mj00021@mix.wvu.edu)). and techniques for incorporating underlying physical principles of the power grid into different types of ML approaches (supervised/semi-supervised learning, unsupervised learning, and reinforcement learning (RL)); b) explore the existing works on PIML methods for anomaly detection, classification, localization, and mitigation in power transmission and distribution systems, c) discuss improvements in existing methods through consideration of potential challenges whi

FAQs

What PIML tasks for power transmission and distribution systems are discussed?

The document focuses on anomaly detection, classification, localization, and mitigation, along with improvements to handle challenges and limitations for real-world application.

Aug 5, 2026

Upload by Finn