본문 바로가기
  • Home

PV Power Forecasting Model for ESS Charge-Discharge

  • Journal of Software Forensics
  • Abbr : JSF
  • 2026, 22(3), pp.133~139
  • Publisher : Korea Software Assessment and Valuation Society
  • Research Area : Engineering > Computer Science
  • Received : August 15, 2026
  • Accepted : September 20, 2026
  • Published : September 30, 2026

Lee Bong Kyu 1

1제주대학교

Accredited

ABSTRACT

Operating an Energy Storage System (ESS) requires scheduling the charging and discharging times. For effective scheduling of photovoltaic (PV)-coupled ESS, it is necessary to forecast PV power generation in advance and use the forecasts for charge–discharge decision-making. However, accurate PV power forecasting is challenging because PV output varies substantially with meteorological conditions and changes in solar position. In this study, we propose a LightGBM-based multi-class classification model that predicts PV generation levels using weather forecasts and historical generation data. The proposed model outputs three generation-level categories—low, medium, and high—normalized by installed capacity. The model is trained and evaluated using data collected from 32 PV facilities installed on Jeju Island.

Journal Copyright Policy

No CCL information provided

Citation status

* References for papers published after 2025 are currently being built.