A comparative study of the stochastic models and harmonically coupled stochastic models in the analysis and forecasting of solar radiation data

摘要: Extra-terrestrially, there is no stochasticity in the solar irradiance, hence deterministic models are often used to model this data. At ground level, Box-Jenkins Seasonal/Non-seasonal Autoregressive Integrated Moving Average (S/ARIMA) short memory stochastic have been such data with some degree of success. This success attributable its ability capture component irradiance series due effects ever-changing atmospheric conditions. However, recorded at earth’s surface rarely entirely but a mixture both and components. One plausible modelling procedure couple sinusoidal predictors determined harmonic (Fourier) frequencies inherent periodicities (seasonalities) diurnal cycle, SARIMA capturing We construct which we term, harmonically coupled (HCSARIMA) use them empirically global horizontal (GHI) surface. Comparison two classes shows that HCSARIMA generally out-compete forecasting arena.