A comparison of the performance of some extreme learning machine empirical models for predicting daily horizontal diffuse solar radiation in a region of southern Iran

Precise knowledge of direct and diffuse solar radiation is important for energy utilization and agricultural activities. However, field measurements in most areas of the world are only for total solar radiation. The satellite-retrieved direct and diffuse solar radiation show poor performance under overcast skies. Therefore, better empirical models are needed to estimate direct and diffuse solar radiation by considering the impact of aerosols over polluted regions. A case study is conducted in North China with the ground-measured solar radiation and satellite-retrieved aerosol optical depth to improve new empirical models at monthly (from 2000 to 2016) and daily (from 2006 to 2009) level. The improved empirical models are validated using the field measurements and compared with the existing models. Results suggest that these models perform well in estimating direct solar radiation at monthly ( R2 = 0.86–0.91, RMSE = 0.76–0.83 MJ/m2) and daily ( R2 = 0.91–0.94, RMSE = 1.51–1.64 MJ/m2) level. The accuracy of estimated monthly ( R2 = 0.95–0.96, RMSE = 0.57–0.65 MJ/m2) and daily ( R2 = 0.91–0.93, RMSE = 1.09–1.15 MJ/m2) diffuse solar radiation, particularly the maximum diffuse solar radiation value, has been improved compared to the existing models. The models presented in this study can be useful in the improvement and evaluation of solar radiation dataset over polluted regions similar to North China.

Download Full-text

Global Solar Radiation Prediction Using Hybrid Online Sequential Extreme Learning Machine Model

Energies ◽

10.3390/en11123415 ◽

2018 ◽

Vol 11 (12) ◽

pp. 3415 ◽

Cited By ~ 10

Author(s):

Muzhou Hou ◽

Tianle Zhang ◽

Futian Weng ◽

Mumtaz Ali ◽

Nadhir Al-Ansari ◽

...

Keyword(s):

Solar Radiation ◽

Extreme Learning Machine ◽

Prediction Accuracy ◽

Prediction Models ◽

Renewable Energy Sources ◽

Information Criterion ◽

Absolute Error ◽

Global Solar Radiation ◽

Machine Model ◽

Learning Machine

Accurate global solar radiation prediction is highly essential for related research on renewable energy sources. The cost implication and measurement expertise of global solar radiation emphasize that intelligence prediction models need to be applied. On the basis of long-term measured daily solar radiation data, this study uses a novel regularized online sequential extreme learning machine, integrated with variable forgetting factor (FOS-ELM), to predict global solar radiation at Bur Dedougou, in the Burkina Faso region. Bayesian Information Criterion (BIC) is applied to build the seven input combinations based on speed (Wspeed), maximum and minimum temperature (Tmax and Tmin), maximum and minimum humidity (Hmax and Hmin), evaporation (Eo) and vapor pressure deficiency (VPD). For the difference input parameters magnitudes, seven models were developed and evaluated for the optimal input combination. Various statistical indicators were computed for the prediction accuracy examination. The experimental results of the applied FOS-ELM model demonstrated a reliable prediction accuracy against the classical extreme learning machine (ELM) model for daily global solar radiation simulation. In fact, compared to classical ELM, the FOS-ELM model reported an enhancement in the root mean square error (RMSE) and mean absolute error (MAE) by (68.8–79.8%). In summary, the results clearly confirm the effectiveness of the FOS-ELM model, owing to the fixed internal tuning parameters.

Download Full-text