View in EDS HTML Full Text PDF Full Text

Behind-the-Meter PV Disaggregation Under Limited Sample Budget: A User-Level Active Learning Strategy.

Saved in:
Bibliographic Details
Title: Behind-the-Meter PV Disaggregation Under Limited Sample Budget: A User-Level Active Learning Strategy.
Authors: Cao, Jiaxu1 (AUTHOR), Chen, Haiwen1,2 (AUTHOR) haiwen.c@hebut.edu.cn, Zhao, Liyuan1,3 (AUTHOR), Wang, Yuzhen2,4 (AUTHOR), Wang, Shaoying1,2 (AUTHOR), Lu, Yanyan2 (AUTHOR), Wang, Jingzhi3,4 (AUTHOR), Lu, Haonan3,4 (AUTHOR)
Source: Energies (19961073). Sep2026, Vol. 19 Issue 18, p4371. 26p.
Subject Terms: *Active learning, *Epistemic uncertainty, *Measurement uncertainty (Statistics), *Signal separation, *Photovoltaic power generation
Abstract: Accurate estimation of Behind-the-meter Photovoltaic (BTM PV) generation is essential for load forecasting and grid planning. Most distributed PV systems are installed behind customer meters, making their output unobservable. Disaggregating PV output from net load is therefore critical for improving distribution network observability. However, existing deep-learning-based disaggregation methods require large labeled datasets, and obtaining such data is costly. Under limited budgets, only a few users can be labeled, which constrains model performance. This paper proposes a user-level BTM PV disaggregation method based on active learning with adaptive weight updates, aiming to maximize model performance with minimal labeling cost. We design a multi-dimensional user value assessment system incorporating epistemic uncertainty, aleatoric uncertainty, and representativeness. In each iteration, the most informative users are selected for sub-meter installation. To dynamically optimize the selection strategy, we propose an adaptive weight update mechanism that adjusts the weights for the next round based on performance improvement gradients across dimensions. This closed-loop feedback enables the strategy to capture evolving model needs and prioritize users that yield the greatest performance gains. The proposed method is validated on the public Ausgrid dataset, and experimental results demonstrate its effectiveness under limited budgets. [ABSTRACT FROM AUTHOR]
Database: Energy & Power Source
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
Text:
  Availability: 1
Header DbId: enr
DbLabel: Energy & Power Source
An: 197391738
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Behind-the-Meter PV Disaggregation Under Limited Sample Budget: A User-Level Active Learning Strategy.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Cao%2C+Jiaxu%22">Cao, Jiaxu</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Chen%2C+Haiwen%22">Chen, Haiwen</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<i> haiwen.c@hebut.edu.cn</i><br /><searchLink fieldCode="AR" term="%22Zhao%2C+Liyuan%22">Zhao, Liyuan</searchLink><relatesTo>1,3</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wang%2C+Yuzhen%22">Wang, Yuzhen</searchLink><relatesTo>2,4</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wang%2C+Shaoying%22">Wang, Shaoying</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Lu%2C+Yanyan%22">Lu, Yanyan</searchLink><relatesTo>2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wang%2C+Jingzhi%22">Wang, Jingzhi</searchLink><relatesTo>3,4</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Lu%2C+Haonan%22">Lu, Haonan</searchLink><relatesTo>3,4</relatesTo> (AUTHOR)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22Energies+%2819961073%29%22">Energies (19961073)</searchLink>. Sep2026, Vol. 19 Issue 18, p4371. 26p.
– Name: Subject
  Label: Subject Terms
  Group: Su
  Data: *<searchLink fieldCode="DE" term="%22Active+learning%22">Active learning</searchLink><br />*<searchLink fieldCode="DE" term="%22Epistemic+uncertainty%22">Epistemic uncertainty</searchLink><br />*<searchLink fieldCode="DE" term="%22Measurement+uncertainty+%28Statistics%29%22">Measurement uncertainty (Statistics)</searchLink><br />*<searchLink fieldCode="DE" term="%22Signal+separation%22">Signal separation</searchLink><br />*<searchLink fieldCode="DE" term="%22Photovoltaic+power+generation%22">Photovoltaic power generation</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Accurate estimation of Behind-the-meter Photovoltaic (BTM PV) generation is essential for load forecasting and grid planning. Most distributed PV systems are installed behind customer meters, making their output unobservable. Disaggregating PV output from net load is therefore critical for improving distribution network observability. However, existing deep-learning-based disaggregation methods require large labeled datasets, and obtaining such data is costly. Under limited budgets, only a few users can be labeled, which constrains model performance. This paper proposes a user-level BTM PV disaggregation method based on active learning with adaptive weight updates, aiming to maximize model performance with minimal labeling cost. We design a multi-dimensional user value assessment system incorporating epistemic uncertainty, aleatoric uncertainty, and representativeness. In each iteration, the most informative users are selected for sub-meter installation. To dynamically optimize the selection strategy, we propose an adaptive weight update mechanism that adjusts the weights for the next round based on performance improvement gradients across dimensions. This closed-loop feedback enables the strategy to capture evolving model needs and prioritize users that yield the greatest performance gains. The proposed method is validated on the public Ausgrid dataset, and experimental results demonstrate its effectiveness under limited budgets. [ABSTRACT FROM AUTHOR]
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=enr&AN=197391738
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.3390/en19184371
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 26
        StartPage: 4371
    Subjects:
      – SubjectFull: Active learning
        Type: general
      – SubjectFull: Epistemic uncertainty
        Type: general
      – SubjectFull: Measurement uncertainty (Statistics)
        Type: general
      – SubjectFull: Signal separation
        Type: general
      – SubjectFull: Photovoltaic power generation
        Type: general
    Titles:
      – TitleFull: Behind-the-Meter PV Disaggregation Under Limited Sample Budget: A User-Level Active Learning Strategy.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Cao, Jiaxu
      – PersonEntity:
          Name:
            NameFull: Chen, Haiwen
      – PersonEntity:
          Name:
            NameFull: Zhao, Liyuan
      – PersonEntity:
          Name:
            NameFull: Wang, Yuzhen
      – PersonEntity:
          Name:
            NameFull: Wang, Shaoying
      – PersonEntity:
          Name:
            NameFull: Lu, Yanyan
      – PersonEntity:
          Name:
            NameFull: Wang, Jingzhi
      – PersonEntity:
          Name:
            NameFull: Lu, Haonan
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 15
              M: 09
              Text: Sep2026
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-print
              Value: 19961073
          Numbering:
            – Type: volume
              Value: 19
            – Type: issue
              Value: 18
          Titles:
            – TitleFull: Energies (19961073)
              Type: main
ResultId 1