Photovoltaic Power Forecasting using LSTM on Limited Dataset

Vidisha De, T. T. Teo, W. L. Woo, T. Logenthiran

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

8 Citations (Scopus)

Abstract

This paper aims to forecast the photovoltaic power, which is beneficial for grid planning which aids in anticipating and prediction in the event of a shortage. Forecasting of photovoltaic power using Recurrent Neural Network (RNN) is the focus of this paper. The training algorithm used for RNN is Long Short-Term Memory (LSTM). To ensure that the amount of energy being harvested from the solar panel is sufficient to match the demand, forecasting its output power will aid to anticipate and predict at times of a shortage. However, due to the intermittent nature of photovoltaic, accurate photovoltaic power forecasting can be difficult. Therefore, the purpose of this paper is to use long short-term memory to obtain an accurate forecast of photovoltaic power. In this paper, Python with Keras is used to implement the neural network model. Simulation studies were carried out on the developed model and simulation results show that the proposed model can forecast photovoltaic power with high accuracy.
Original languageEnglish
Title of host publicationInternational Conference on Innovative Smart Grid Technologies, ISGT Asia 2018
PublisherIEEE
Pages710-715
Number of pages6
ISBN (Electronic)9781538642917, 9781538642900
ISBN (Print)9781538642924
DOIs
Publication statusPublished - 20 Sep 2018
Event2018 IEEE Innovative Smart Grid Technologies - Asia - Suntec Singapore International Convention and Exhibition Centre, Singapore, Singapore
Duration: 22 May 201825 May 2018
http://sites.ieee.org/isgt-asia-2018/

Conference

Conference2018 IEEE Innovative Smart Grid Technologies - Asia
Abbreviated titleISGT Asia 2018
CountrySingapore
CitySingapore
Period22/05/1825/05/18
Internet address

Fingerprint Dive into the research topics of 'Photovoltaic Power Forecasting using LSTM on Limited Dataset'. Together they form a unique fingerprint.

Cite this