Scholarlink

<p>Smart inverters with grid support functions are commonly used in distributed solar photovoltaic systems to prevent overvoltage by curtailing energy. However, estimating energy curtailment is challenging for utilities due to limited access to customers’ inverter settings and the lack of site-specific climate data. This paper proposes a novel AMI-only method that remotely detects Volt-Watt settings via Hough transform-based inference and estimates long-term energy curtailment using historical voltage and active power measurements. Year-long quasi-static time-series (QSTS) simulations on a benchmark feeder show annual curtailment estimation errors below 14.6% for the curtailed systems (RMSE &lt; 0.026&#xa0;p.u., <i>R</i><sup>2</sup> up to 0.748) and improved accuracy versus an AMI-only reference method. A field study in Hong Kong further confirms correct Volt-Watt detection and reliable annual curtailment estimation for events observed on 12&#xa0;days. The proposed method provides utilities with a scalable and practical way to infer inverter behaviour and estimate long-term curtailment without additional communications, network models, or climate sensors.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

AMI data-based remote detection of PV inverter grid support function and energy curtailment estimation

  • 期刊论文

  • 发表日期：

    2026年2月23日

查看全文
  • Ang Li,
  • C. Y. Chan,
  • Edmund Fung,
  • Hillman Lai,
  • W. H. Yim,
  • Cady Mak,
  • Ricky Tam,
  • Simon Ng

摘要

Smart inverters with grid support functions are commonly used in distributed solar photovoltaic systems to prevent overvoltage by curtailing energy. However, estimating energy curtailment is challenging for utilities due to limited access to customers’ inverter settings and the lack of site-specific climate data. This paper proposes a novel AMI-only method that remotely detects Volt-Watt settings via Hough transform-based inference and estimates long-term energy curtailment using historical voltage and active power measurements. Year-long quasi-static time-series (QSTS) simulations on a benchmark feeder show annual curtailment estimation errors below 14.6% for the curtailed systems (RMSE < 0.026 p.u., R2 up to 0.748) and improved accuracy versus an AMI-only reference method. A field study in Hong Kong further confirms correct Volt-Watt detection and reliable annual curtailment estimation for events observed on 12 days. The proposed method provides utilities with a scalable and practical way to infer inverter behaviour and estimate long-term curtailment without additional communications, network models, or climate sensors.