Short-term forecasting for solar irradiation based on the multi-layer neural network with the Levenberg-Marquardt algorithm and meteorological data: application to the Gandon site in Senegal

Nkounga, Willy Magloire; Ndiaye, Mouhamadou Falilou; Ndiaye, Mamadou Lamine; Cisse, Oumar Ibn Khatab; Bop, Mamadou; Alexandre, Sioutas

doi:10.1109/icrera.2018.8566850

Cited by 11 publications

(6 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…More than anything else, the accuracy of ANN models depend on location and weather predictability. The models that had lower nRMSE values tended to be in locations such as Greece [15], Corsica island [16] and Senegal [17] where the weather is consistently sunny for a significant portion of the year. At 10.7% and 10.8% nRMSE, the error of the single-layer models presented in this article fall at the low end of the published spectrum.…”

Section: Discussionmentioning

confidence: 99%

Training and Testing of a Single-Layer LSTM Network for Near-Future Solar Forecasting

et al. 2020

View full text Add to dashboard Cite

Increasing integration of renewable energy sources, like solar photovoltaic (PV), necessitates the development of power forecasting tools to predict power fluctuations caused by weather. With trustworthy and accurate solar power forecasting models, grid operators could easily determine when other dispatchable sources of backup power may be needed to account for fluctuations in PV power plants. Additionally, PV customers and designers would feel secure knowing how much energy to expect from their PV systems on an hourly, daily, monthly, or yearly basis. The PROGNOSIS project, based at the Cyprus University of Technology, is developing a tool for intra-hour solar irradiance forecasting. This article presents the design, training, and testing of a single-layer long-short-term-memory (LSTM) artificial neural network for intra-hour power forecasting of a single PV system in Cyprus. Four years of PV data were used for training and testing the model (80% for training and 20% for testing). With a normalized root mean squared error (nRMSE) of 10.7%, the single-layer network performed similarly to a more complex 5-layer LSTM network trained and tested using the same data. Overall, these results suggest that simple LSTM networks can be just as effective as more complicated ones.

show abstract

Section: Discussionmentioning

confidence: 99%

Training and Testing of a Single-Layer LSTM Network for Near-Future Solar Forecasting

et al. 2020

View full text Add to dashboard Cite

show abstract

“…The meteorological data used are measured and averaged in 10 min steps; they are: number of the day (Dn), time of measurement (h), maximum (Tm) and average (T) temperature, maximum (Hm) and average (H) relative humidity, (Em) and average (E) solar irradiation, atmospheric pressure (P). These data are presented and detailed in our previous work [15].…”

Section: Study Environment and Data Presentationmentioning

confidence: 98%

“…The ANN typology used is the feed-forward multilayer perceptron neural network. Its internal configuration as well as the training equations (the levenbergmarquardt agorithm) are presented in [15]. The ANN architecture and the training flowchart for each horizon are presented respectively on Figure 2 ANN performance are calculated using Eq.…”

Section: Artificial Neural Networkmentioning

confidence: 99%

“…The works [14] has developed a short-term forecasting model for wind potential in the north-west of Senegal using linear prediction, with a correlation coefficient of 0.86. Solar irradiation short-term forecasting, applied to north-western Senegal is developed in [15] with a correlation coefficient of 0.99; the authors have limited the horizon to 10 min. To complete this limit, we propose forecasting models for short-term horizons from 30 minutes to 06 hours.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Short-term Multi Horizons Forecasting of Solar Irradiation Based on Artificial Neural Network with Meteorological Data: Application in the North-west of Senegal

Nkounga

Ndiaye

Cisse

et al. 2021

2021 Sixteenth International Conference on Ecological Vehicles and Renewable Energies (EVER)

Self Cite

View full text Add to dashboard Cite

This study proposes a short term forecasting of solar irradiation with multi horizons in the north-west of Senegal. The multilayer artificial neural network (ANN), based on the Levenberg Marquardt algorithm and the meteorological data are used. The latter are measured in real time on the study site. The variables of interest are: mean solar irradiation, maximum temperature and measurement time; they are selected using Weka software. The forecasting horizons are: 0.5 hour, 1 hour, 1.5 hours, 2 hours, 2.5 hours, 3 hours, 3.5 hours, 4 hours, 4.5 hours, 5 hours, 5.5 hours and 06 hours. They are proposed with the corresponding statistical criteria. The results show that, the solar energy forecasting can be extended over a six-hour horizon with a correlation coefficient of 0.97 and root mean square error of 0.07. These results will make it possible to complete the forecasting tools in the solar energy sector in Senegal, and help investors to choose the most suitable horizons for energy forecasting in photovoltaic solar power plants.

show abstract

“…In this paper, ANN is used. Its interest has been presented in our previous work [33]. It will allow to quantify the available photovoltaic power over the control horizon.…”

Section: Introductionmentioning

confidence: 99%

Automatic control and dispatching of charging currents to a charging station for power-assisted bikes

Nkounga

Ndiaye

Cisse

et al. 2022

Energy

Self Cite

View full text Add to dashboard Cite

Short-term forecasting for solar irradiation based on the multi-layer neural network with the Levenberg-Marquardt algorithm and meteorological data: application to the Gandon site in Senegal

Cited by 11 publications

References 26 publications

Training and Testing of a Single-Layer LSTM Network for Near-Future Solar Forecasting

Training and Testing of a Single-Layer LSTM Network for Near-Future Solar Forecasting

Short-term Multi Horizons Forecasting of Solar Irradiation Based on Artificial Neural Network with Meteorological Data: Application in the North-west of Senegal

Automatic control and dispatching of charging currents to a charging station for power-assisted bikes

Contact Info

Product

Resources

About