Deep Learning Technique for Forecasting Solar Radiation and Wind Speed for Dynamic Microgrid Analysis

Md Mainul Islam, Hussain Shareef, Eslam Salah Fayez Al Hassan

Research output: Contribution to journalArticlepeer-review

Abstract

The key variables in the development and operation of wind and solar power systems are wind speed and solar radiation. The prediction of solar and wind energy parameters is important to alleviate the effects of power generation fluctuations. Consequently, it is essential to predict renewable energy sources like solar radiation and wind speed precisely. An artificial intelligence-based random forest method is recommended in this paper to estimate wind speed and solar radiation. The number of decision trees in the random forest model is suggested to be optimised using a novel coot algorithm (CA), and the effectiveness of the CA is evaluated to that of the currently used particle swarm optimisation (PSO) method. The best forecasting data are used in this work to develop a dynamic Microgrid (MG) in MATLAB/SIMULINK. A novel binary CA is proposed to control the MG to minimize the cost. The effect of the energy storage system is also investigated during the simulation of the MG.

Original languageEnglish
Pages (from-to)162-170
Number of pages9
JournalPrzeglad Elektrotechniczny
Volume99
Issue number4
DOIs
Publication statusPublished - 2023

Keywords

  • Solar power
  • coot algorithm
  • forecasting
  • microgrid
  • random forest method
  • wind power

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Deep Learning Technique for Forecasting Solar Radiation and Wind Speed for Dynamic Microgrid Analysis'. Together they form a unique fingerprint.

Cite this