Abstract
The power grid is undergoing massive changes to ensure resiliency and reliability in a more decentralized world. Distributed energy resources are becoming a prominent source of generation, potentially leading to a lack of centralized generation sources. Due to these new behaviors and system topologies, it is important to install measurement devices that are 1) accurate and 2) self-aware of their measurement quality. In this paper, a residential-scale microgrid is used to generate voltage and current waveforms, captured by Verivolt and National Instruments measurement equipment. A least-squares approach is used to separate the "clean"signals from the noise. Gaussian mixture modeling is used to approximate noise distributions, and it is shown these higher-order distribution estimates are a better fit to voltage and current noise profiles than single-mode Gaussian estimates.
| Original language | English |
|---|---|
| Pages (from-to) | 3547-3550 |
| Number of pages | 4 |
| Journal | IEEE Transactions on Power Delivery |
| Volume | 39 |
| Issue number | 6 |
| DOIs | |
| State | Published - 2024 |
Funding
This work was supported in part by the US Department of Energy (DOE), Office of Electricity, under Contract DE-AC05-00OR22725 with UT-Battelle, LLC, for the US DOE, and in part by the AI Initiative Project funded by Oak Ridge National Laboratory.
Keywords
- Gaussian mixture modeling
- noise estimation
- power system measurement
Fingerprint
Dive into the research topics of 'Statistical Behavior of Low-Amplitude Power System Point-on-Wave Measurements'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver