Data Based Solar Radiation Modelling - Hydrological Data Driven Modelling - page 131

Fig. 5.28 Line diagram showing solar radiation as observed and estimated using the

Levenberg Marquardt ANN in training phase

-

Fig. 5.29 Line diagram showing solar radiation as observed and estimated using the

Levenberg - Marquardt ANN in testing phase

5.4.2.2 Daily Solar Radiation Modelling with SVM and ANFIS Models

The adaptive neural-based fuzzy inference system (ANFIS) model was used for
daily solar radiation modelling, in which a set of parameters in ANFIS identi ed

through a hybrid learning rule combining the back-propagation gradient descent and
a least squares method. The ANFIS model in solar radiation modelling has used a set
of fuzzy IF

THEN rules with appropriate membership functions. The mathematical
details of the ANFIS function are given in the Chap. 4 . The ANFIS model has used

-

first 770 data points of the whole available data points as training data set as per the
recommendations of the Gamma Test. The remaining data points were used for the
testing stage. The input layer of the ANFIS consisted of data series like horizontal
extraterrestrial radiation, daily mean air temperature (averaged over 24 h), maximum
daily air temperature, minimum daily air temperature, daily wind speed (averaged