An official website of the United States government
Estimated component failure interactions from utility outage data can capture the general failure propagation patterns and help identify key components of a power system. Conventionally, utility outages are grouped into cascades and generations according to inter-outage time based on arbitrarily chosen thresholds. In this paper, we propose an optimal decomposition approach for utility outage data. By approximating the temporal pattern of the outage sequence by a Poisson process, an adaptive generation duration threshold is calculated to group the generations for each cascade. Compared to the conventional method, the proposed method can reveal more failure interactions and mitigate heterogeneity. The results based on real utility outage data demonstrate the effectiveness of the proposed optimal decomposition approach.
more »
« less
Learning Cascading Failure Interactions by Deep Convolutional Generative Adversarial Network
In this paper, a cascading failure interaction learning method is proposed for real utility outage data. For better revealing the structure, we reorganize the failure interaction matrix based on Louvain community detection. A deep convolutional generative adversarial network (DCGAN) based method is then proposed to learn the implicit features for failure propagation in the interaction matrix. A systematic method is further developed to evaluate the performance of the learning method on missing interaction recovery and new interaction discovery. The effectiveness of the proposed method is validated on the 14-year real utility outage data from Bonneville Power Administration.
more »
« less
- PAR ID:
- 10385081
- Date Published:
- Journal Name:
- 2022 IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids (SmartGridComm)
- Page Range / eLocation ID:
- 21 to 26
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
In recent decades, blackouts have shown an increasing prevalence of power outages due to extreme weather events such as hurricanes. Precisely assessing the spatiotemporal outages in distribution networks, the most vulnerable part of power systems, is critical to enhancing power system resilience. The Sequential Monte Carlo (SMC) simulation method is widely used for spatiotemporal risk analysis of power systems during extreme weather hazards. However, it is found here that the SMC method can lead to large errors as it repeatedly samples the failure probability from the time-invariant fragility functions of system components in time-series analysis, particularly overestimating damages under evolving hazards with high-frequency sampling. To address this issue, a novel hazard resistance-based spatiotemporal risk analysis (HRSRA) method is proposed. This method converts the failure probability of a component into a hazard resistance and uses it as a time-invariant value in time-series analysis. The proposed HRSRA provides an adaptive framework for incorporating high-spatiotemporal-resolution meteorology models into power outage simulations. By leveraging the geographic information system data of the power system and a physics-based hurricane wind field model, the superiority of the proposed method is validated using real-world time-series power outage data from Puerto Rico, including data collected during Hurricane Fiona in 2022.more » « less
-
null (Ed.)Recent hurricane events have caused unprecedented amounts of damage on critical infrastructure systems and have severely threatened our public safety and economic health. The most observable (and severe) impact of these hurricanes is the loss of electric power in many regions, which causes breakdowns in essential public services. Understanding power outages and how they evolve during a hurricane provides insights on how to reduce outages in the future, and how to improve the robustness of the underlying critical infrastructure systems. In this article, we propose a novel scalable segmentation with explanations framework to help experts understand such datasets. Our method, CnR (Cut-n-Reveal), first finds a segmentation of the outage sequences based on the temporal variations of the power outage failure process so as to capture major pattern changes. This temporal segmentation procedure is capable of accounting for both the spatial and temporal correlations of the underlying power outage process. We then propose a novel explanation optimization formulation to find an intuitive explanation of the segmentation such that the explanation highlights the culprit time series of the change in each segment. Through extensive experiments, we show that our method consistently outperforms competitors in multiple real datasets with ground truth. We further study real county-level power outage data from several recent hurricanes (Matthew, Harvey, Irma) and show that CnR recovers important, non-trivial, and actionable patterns for domain experts, whereas baselines typically do not give meaningful results.more » « less
-
This paper presents a variational Bayesian inference Neural Network (BNN) approach to quantify uncertainties in matrix function estimation for the state-space linear parameter-varying (LPV) model identification problem using only inputs/outputs data. The proposed method simultaneously estimates states and posteriors of matrix functions given data. In particular, states are estimated by reaching a consensus between an estimator based on past system trajectory and an estimator by recurrent equations of states; posteriors are approximated by minimizing the Kullback–Leibler (KL) divergence between the parameterized posterior distribution and the true posterior of the LPV model parameters. Furthermore, techniques such as transfer learning are explored in this work to reduce computational cost and prevent convergence failure of Bayesian inference. The proposed data-driven method is validated using experimental data for identification of a control-oriented reactivity controlled compression ignition (RCCI) engine model.more » « less
-
The paper proposes a novel approach for the outage State of Risk (SoR) assessment caused by weather and vegetation in the distribution grid. Machine Learning prediction algorithm is used in conjunction with GIS application for mapping the SoR for the entire network. The proposed optimization approach leads to the specification of the mitigation strategies that utility staff and customers can coordinate to minimize the impact of outages. The resulting SoR assessment enables the implementation of an innovative decision- making solution for utility operators, represented in the form of risk maps. Additionally, utilizing the SoR assessments, a Customer Notification System (CNS) is introduced to enhance customer awareness and facilitate the adoption of mitigation measures. This holistic approach shifts outage management from a reactive process to a proactive initiative, promoting grid resilience and reliability through planned outage mitigation.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1109/SmartGridComm52983.2022.9961045
