Dynamic Anomaly Detection With High-Fidelity Simulators: A Convex Optimization Approach

Abstract

The main objective of this article is to develop scalable dynamic anomaly detectors with high-fidelity simulators of power systems. On the one hand, models in high-fidelity simulators are typically "intractable" if one opts to describe them in a mathematical formulation in order to apply existing modelbased approaches from the anomaly detection literature. On the other hand, pure data-driven methods developed primarily in the machine learning literature neglect our knowledge about the underlying dynamics of power systems. In this study, we combine tools from these two mainstream approaches to develop a data-assisted model-based diagnosis filter utilizing both the knowledge from a picked abstract model and also the data of simulation results from high-fidelity simulators. The proposed diagnosis filter aims to achieve two desired features: (i) performance robustness with respect to model mismatch; (ii) high scalability. To this end, we propose a tractable (convex) optimization-based reformulation in which decisions are the filter parameters, the model-based information introduces feasible sets, and the data from the simulator forms the objective function to-be-minimized regarding the effect of model mismatch on the filter performance. To validate the theoretical results, we implement the developed diagnosis filter in DIgSILENT PowerFactory to detect false data injection attacks on the Automatic Generation Control measurements in the three-area IEEE 39-bus system.

Ultra Local Nonlinear Unknown Input Observers for Robust Fault Reconstruction

“…In this result, by constraining the mapping from the fault to the filter output, the fault signal can be estimated. Similar results can be found in [12] where instead of minimizing the nonlinearity effect, the output mismatch of the actual and simulation-based system is minimized to provide robustifcation against model mismatch.…”

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

“…Ref [109] develops a robust and scalable tool based on a combination of the knowledge of the system provided by an explicit but reduced-order mathematical model (abstract model) and a data-driven approach for dynamic anomaly detection in power systems. In the proposed method, a highfidelity simulator describing the mathematical model of the system is presented to represent the dynamics of the system, and in the second phase, a filter for detecting both multivariate and univariate cyber attacks is employed to minimize the effects of the difference between the outputs of the highfidelity simulator and the abstract model.…”

Multimode diagnosis for switched affine systems with noisy measurement

“…Another method is to develop adaptive nonlinear estimators to approximate the nonlinear terms (Boem, Ferrari, & Parisini, 2011;Ferrari, Parisini, & Polycarpou, 2011). More recently, the authors in Mohajerin Esfahani and Lygeros (2015), Pan, Palensky, and Mohajerin Esfahani (2021) develop tractable optimization-based approaches in the DAE framework to design FDI filters to deal with disturbances and nonlinear terms.…”