AI and ML Applications for PV Reliability and System Performance

Perry, Kirsten (ORCID:0000000280957472); Deline, Chris (ORCID:0000000298678930); Jordan, Dirk (ORCID:0000000221837489); Deceglie, Michael; Muller, Matthew (ORCID:0000000238354096); Springer, Martin (ORCID:000000016803108X); White, Robert (ORCID:0000000252084061); Meyers, B.; Ragsdale, D.; Miskovich, S.; Ogut, G.

DOE OSTI · 2274802

AI and ML Applications for PV Reliability and System Performance

Perry, Kirsten (ORCID:0000000280957472)·Deline, Chris (ORCID:0000000298678930)·Jordan, Dirk (ORCID:0000000221837489)·Deceglie, Michael·Muller, Matthew (ORCID:0000000238354096)·Springer, Martin (ORCID:000000016803108X)·White, Robert (ORCID:0000000252084061)·Meyers, B.·Ragsdale, D.·Miskovich, S.·Ogut, G.

View original source ↗Read source document ↗Export metadata ↓

Abstract

This poster discusses AI and ML topics in PV reliability and system performance. In particular, automated metadata extraction and QA for fielded solar installations is covered for the PV Fleets Project. Additionally, statistical learning topics for the PVInsight Project are addressed, as well as development of the PV Validation Hub.

Explore related subjects

algorithm analytics deep learning machine learning MATHEMATICS AND COMPUTING,SOLAR ENERGY solar

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Perry, Kirsten (ORCID:0000000280957472), Deline, Chris (ORCID:0000000298678930), Jordan, Dirk (ORCID:0000000221837489), Deceglie, Michael, Muller, Matthew (ORCID:0000000238354096), Springer, Martin (ORCID:000000016803108X), White, Robert (ORCID:0000000252084061), Meyers, B., Ragsdale, D., Miskovich, S., Ogut, G.. 2023-12-15. AI and ML Applications for PV Reliability and System Performance. https://www.osti.gov/biblio/2274802

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related reports

DOE OSTI2024

Demonstration of the rodeo algorithm on a quantum computer

The rodeo algorithm is an efficient algorithm for eigenstate preparation and eigenvalue estimation for any observable on a quantum computer. This makes it a promising tool for studying the spectrum and structure of atomic nuclei as well as other fields of quantum many-body physics. The only requirement is that the initial state has sufficient overlap probability with the desired eigenstate. While it is exponentially faster than well-known algorithms such as phase estimation and adiabatic evolution for eigenstate preparation, it has yet to be implemented on an actual quantum device. In this work, we apply the rodeo algorithm to determine the energy levels of a random one-qubit Hamiltonian, resulting in a relative error of 0.08% using mid-circuit measurements on the IBM Q device Casablanca. This surpasses the accuracy of directly-prepared eigenvector expectation values using the same quantum device. We take advantage of the high-accuracy energy determination and use the Hellmann-Feynman theorem to compute eigenvector expectation values for a different random one-qubit observable. For the Hellmann-Feynman calculations, we find a relative error of 0.7%. Here, we conclude by discussing possible future applications of the rodeo algorithm for multi-qubit Hamiltonians.

algorithm↗

DOE OSTI2023

Linearized Distribution Optimal Power Flow for OEDI SI

This research is to meant to demonstrate the OEDI SI use case for distributed optimal power flow (DOPF). The goal was to formulate the optimal power flow problem in the distribution system for active and reactive power setpoints of PV systems using topology information and voltage measurements. The co-simulation runs every 15 minutes as outlined within the scenario file for the given feeder configuration. The linked GitHub repository includes five federates to achieve DOPF for the small, medium, large, and IEEE 123 feeder scenarios. We are using the OEDI SI framework, as well as the example feeder, sensor, recorder, and estimator federates provided in the example repository for OEDI SI. We also provide a runner script for switching between scenarios.

algorithm↗

DOE OSTI2023

PV Validation Hub

The Validation Hub will be a clearinghouse for the transfer of novel algorithms and software from the PV research community to industry. Potential algorithms tested in the Hub could include the estimation of various PV loss factors and the detection of various operational issues. The primary function of the Hub will be for developers to submit executable code which will run on hosted data sets. Developers will receive private reports on the accuracy and performance (e.g., run-time) of the submitted algorithms, and public high level summaries will be hosted. These summaries will indicate the organization who submitted the algorithm (e.g., links to GitHub pages, documentation websites, etc.), high-level accuracy metrics, and standardized performance metrics. These results will be stored in a publicly available database, accessible through the Hub, with the ability for users to sort and filter the results. In short, the Hub will be presented to public users as a collection of interactive leaderboards, organized around specific analysis tasks pertinent to the PV data science community. These tasks include things such as the estimation of various PV loss factors and the detection of various operational issues. We will present progress on the development of this hub, including preliminary results of comparative validation of PV data science algorithms and progress towards building the platform itself.

algorithm↗