Development of a Short Term Solar Power Forecaster Using Artificial Neural Network and Particle Swarm Optimization Techniques (ANN-PSO)

Adeyemi-Kayode, Temitope M. and Orovwode, H. E. and Williams, Chibuzor T. and Adoghe, A. U. and Chouhan, Virendra Singh and Misra, Sanjay (2022) Development of a Short Term Solar Power Forecaster Using Artificial Neural Network and Particle Swarm Optimization Techniques (ANN-PSO). In: International Conference on Machine Intelligence and Signal Processing, 2023, Online.

PDF Download (243kB)

Abstract

Globally, the use of renewable energy has increased significantly since the late twentieth century. Nigeria is also leading the exponential growth of renewable energy use. This article will predict the solar energy collected in 11 power distribution company areas (DISCO) in Nigeria: Abuja, Benin, Eko, Enugu, Ibadan, Ikeja, Jos, Kaduna, Kano, Port-Harcourt, and Yola. Artificial Neural networks and Particle Swarm Optimization (ANN-PSO) techniques are used to forecast solar irradiance. This research compares the results using cognitive acceleration coefficients. From this study, the regression coefficient (R) values of 0.9968 and 0.99533 were obtained from Yola and Ikeja Distribution Company, respectively. Also, mean absolute percentage error (MAPE) values of 3.07% in Yola and 5.67% in Jos were obtained. The normalized root means square error (nRMSE) values of 0.9813, 2.4522, and 0.9470 were obtained from Yola, Ikeja, and Benin DISCOs, respectively, and mean squared error (MSE) values of 2.29% in Abuja, 1.80% in Ibadan, 1.83% in Ikeja, and 0.0915% in Jos. The simulation was also performed for July 2021, which was not part of the dataset used in this study. The result of the forecaster revealed high levels of forecasting accuracy.

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	Solar forecasting  Sustainability  Artificial neural network  Particle swarm optimization  Nigeria
Subjects:	T Technology > T Technology (General) T Technology > TK Electrical engineering. Electronics Nuclear engineering
Divisions:	Faculty of Engineering, Science and Mathematics > School of Electronics and Computer Science Faculty of Engineering, Science and Mathematics > School of Engineering Sciences
Depositing User:	nwokealisi
Date Deposited:	04 Aug 2023 08:25
Last Modified:	04 Aug 2023 08:25
URI:	http://eprints.covenantuniversity.edu.ng/id/eprint/17237

Actions (login required)

View Item