Browsing by Author "Boahen, Samuel"

Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    An Assessment of the impacts of selected Meteorological and Land Use Land Cover Datasets on the accuracy of wind speeds downscaled with the Weather Research and Forecasting Model for coastal areas in Ghana
    (Journal of Renewable Energies, 2024-06-27) Dzebre, Denis Edem Kwame; Asiedu, Charlotte; Akowuah, Eric; Boahen, Samuel; Amoabeng, Kofi Owura; Oppong, David
    Downscaling of wind speeds with the Weather Research and Forecasting model (WRF) model requires inputs from datasets such as Meteorological and Land Use and Land Cover (LULC) datasets. The accuracy of these datasets is among the factors that significantly impact the accuracy of the wind speeds that are generated by the model. In this study, we assess the accuracy of wind speeds data that are downscaled for an area in coastal Ghana using six meteorological, and two global Land use and Land Cover (LULC) datasets as inputs to the WRF model. In contrast to the LULC datasets tested, model wind speeds for the area were more significantly impacted by the different meteorological datasets. Meteorological datasets that were produced with higher resolution forecasts combined with more advanced data assimilation techniques produced better estimates of wind speed, and vice versa. The JMA JRA55 Reanalysis, NCEP GFS Analysis data, and ECWMF ERA5 gave the relatively best combinations of wind speed error metrics and are therefore recommended for consideration for downscaling of wind speeds for wind resources assessment in the coastal regions of Ghana. However, the ECWMF ERA5 is preferred as its mean error margins are fairly constant and so should be easier to correct.
  • No Thumbnail Available
    Item
    Machine learning forecasting of solar PV production using single and hybrid models over different time horizons
    (Heliyon, 2024-04-15) Asiedu, Shadrack T.; Nyarko, Frank Kwabena Afriyie; Boahen, Samuel; Effah, Francis Boafo; Asaaga, Benjamin Atribawuni
    This study uses operational data from a 180 kWp grid-connected solar PV system to train and compare the performance of single and hybrid machine learning models in predicting solar PV production a day-ahead, a week-ahead, two weeks ahead and one month-ahead. The study also analyses the trend in solar PV production and the effect of temperature on solar PV production. The performance of the models is evaluated using R 2 score, mean absolute error and root mean square error. The findings revealed the best-performing model for the day ahead forecast to be Artificial Neural Network. Random Forest gave the best performance for the two-week and a month-ahead forecast, while a hybrid model composed of XGBoost and Random Forest gave the best performance for the week-ahead prediction. The study also observed a downward trend in solar PV production, with an average monthly decline of 244.37 kWh. Further, it was observed that an increase in the module temperature and ambient temperature beyond 47 ◦ C and 25 ◦ C resulted in a decline in solar PV production. The study shows that machine learning models perform differently under different time horizons. Therefore, selecting suitable machine learning models for solar PV forecasts for varying time horizons is extremely necessary.