A Hybrid Data-Driven and Model-Based Anomaly Detection Scheme for DER Operation
This paper proposes a hybrid data and model-based anomaly detection scheme to secure the operation of distributed energy resources (DERs) in distribution grids. Data-driven autoencoders are set up at the edge device level and they use local DER operational data as inputs. The abnormal statuses are detected by analyzing reconstruction errors. In parallel, modelbased state estimation (SE) is set up at the central level and it uses system-wide models and measurements as data inputs. The anomalies are identified by analyzing measurement residuals. The hybrid scheme preserves the benefits of both data-driven and model-based analyses and thus improves the robustness and the accuracy of anomaly detection. Numerical tests based on the model of a real distribution feeder in Southern California highlight the proposed scheme's effectiveness and benefits.
- Research Organization:
- National Renewable Energy Laboratory (NREL), Golden, CO (United States)
- Sponsoring Organization:
- USDOE Grid Modernization Laboratory Consortium
- DOE Contract Number:
- AC36-08GO28308
- OSTI ID:
- 1883207
- Report Number(s):
- NREL/CP-5D00-83787; MainId:84560; UUID:021f4c8e-c3df-4996-b521-4d41950aa273; MainAdminID:65157
- Country of Publication:
- United States
- Language:
- English
Similar Records
Generative and Encoded Anomaly Detectors
Self-Taught Anomaly Detection With Hybrid Unsupervised/Supervised Machine Learning in Optical Networks