A Case Study of Using Long Short-Term Memory (LSTM) Algorithm in Solar Photovoltaic Power Forecasting

Lee Chin, Kho and Sze Song, Ngu and Annie, Joseph and Siti Kudnie, Sahari and Kuryati, Kipli and R., Rulaningtyas (2023) A Case Study of Using Long Short-Term Memory (LSTM) Algorithm in Solar Photovoltaic Power Forecasting. ASM Science Journal, 18 (2023). pp. 1-8. ISSN 1823-6782

PDF 60 long_short_term-1-2.pdf Download (172kB)

Abstract

Solar photovoltaic power plays an important role in distributed energy resources. The number of solar-powered electricity generation has increased steadily in recent years all over the world. This happens because it produces clean energy, and solar photovoltaic technology is continuously developing. One of the challenges in solar photovoltaic is that power generation is highly dependent on the dynamic changes of environmental parameters and asset operating conditions. Solar power forecasting can be a possible solution to maximise the electricity generation capability of the solar photovoltaic system. This study implements the deep learning method, long short-term memory (LSTM) models for time series forecasting in solar photovoltaic power generation forecasting. The data set collected by The Ravina Project from 2010 to 2014 is used as the training data in the simulations. The root mean square value is used in this study to measure the forecasting error. The results show that the deep learning algorithm provides reliable forecasting results.

Actions (For repository members only: login required)

View Item