Hourly irradiance forecasting in Malaysia using support vector machine

摘要: This paper investigates the use of support vector machine (SVM) to forecast hourly solar irradiance for a tropical country. The data was obtained from Sepang Malaysia, recorded throughout 2011. is converted into corresponding clearness index values facilitate model convergence. tested against standard multilayer perceptron (MLP) technique and persistence forecast. evaluation metrics used validate each model's performance are mean bias error, root square absolute error/average, Kolmogorov- Smirnov integral test. Results show that SVM performs significantly better than conventional MLP technique.