IWA Publishing

Table 3

Results of combining models for both case studies for four seasons

Case study	Season	Combining model	Model structure^a	DC		RMSE^b
Case study	Season	Combining model	Model structure^a	Calibration	Verification	Calibration	Verification
Tabriz	Spring	Simple averaging	–	0.91	0.81	0.068	0.054
		Weighted averaging	0.342-0.333-0.325	0.92	0.82	0.066	0.050
		Neural averaging	3-5-1	0.95	0.89	0.014	0.018
	Summer	Simple averaging	–	0.87	0.79	0.071	0.066
		Weighted averaging	0.346-0.335-0.319	0.87	0.81	0.070	0.064
		Neural averaging	3-6-1	0.93	0.89	0.038	0.045
	Autumn	Simple averaging	–	0.87	0.78	0.073	0.060
		Weighted averaging	0.346-0.338-0.316	0.88	0.80	0.072	0.060
		Neural averaging	3-4-1	0.92	0.90	0.041	0.047
	Winter	Simple averaging	–	0.83	0.71	0.091	0.0103
		Weighted averaging	0.343-0.333-0.324	0.85	0.72	0.091	0.101
		Neural averaging	3-8-1	0.91	0.86	0.057	0.062
Rasht	Spring	Simple averaging	–	0.84	0.76	0.074	0.055
		Weighted averaging	0.341-0.332-0.327	0.84	0.77	0.073	0.053
		Neural averaging		0.94	0.90	0.051	0.0.34
	Summer	Simple averaging	–	0.84	0.75	0.086	0.062
		Weighted averaging	0.348-0.337-0.315	0.85	0.76	0.084	0.061
		Neural averaging		0.91	0.90	0.046	0.047
	Autumn	Simple averaging	–	0.80	0.75	0.088	0.067
		Weighted averaging	0.348-0.338-0.314	0.81	0.75	0.086	0.067
		Neural averaging		0.91	0.87	0.052	0.049
	Winter	Simple averaging	–	0.75	0.70	0.099	0.110
		Weighted averaging	0.346-0.332-0.322	0.77	0.73	0.097	0.108
		Neural averaging		0.90	0.86	0.061	0.069

Case study	Season	Combining model	Model structure^a	DC		RMSE^b
Case study	Season	Combining model	Model structure^a	Calibration	Verification	Calibration	Verification
Tabriz	Spring	Simple averaging	–	0.91	0.81	0.068	0.054
		Weighted averaging	0.342-0.333-0.325	0.92	0.82	0.066	0.050
		Neural averaging	3-5-1	0.95	0.89	0.014	0.018
	Summer	Simple averaging	–	0.87	0.79	0.071	0.066
		Weighted averaging	0.346-0.335-0.319	0.87	0.81	0.070	0.064
		Neural averaging	3-6-1	0.93	0.89	0.038	0.045
	Autumn	Simple averaging	–	0.87	0.78	0.073	0.060
		Weighted averaging	0.346-0.338-0.316	0.88	0.80	0.072	0.060
		Neural averaging	3-4-1	0.92	0.90	0.041	0.047
	Winter	Simple averaging	–	0.83	0.71	0.091	0.0103
		Weighted averaging	0.343-0.333-0.324	0.85	0.72	0.091	0.101
		Neural averaging	3-8-1	0.91	0.86	0.057	0.062
Rasht	Spring	Simple averaging	–	0.84	0.76	0.074	0.055
		Weighted averaging	0.341-0.332-0.327	0.84	0.77	0.073	0.053
		Neural averaging		0.94	0.90	0.051	0.0.34
	Summer	Simple averaging	–	0.84	0.75	0.086	0.062
		Weighted averaging	0.348-0.337-0.315	0.85	0.76	0.084	0.061
		Neural averaging		0.91	0.90	0.046	0.047
	Autumn	Simple averaging	–	0.80	0.75	0.088	0.067
		Weighted averaging	0.348-0.338-0.314	0.81	0.75	0.086	0.067
		Neural averaging		0.91	0.87	0.052	0.049
	Winter	Simple averaging	–	0.75	0.70	0.099	0.110
		Weighted averaging	0.346-0.332-0.322	0.77	0.73	0.097	0.108
		Neural averaging		0.90	0.86	0.061	0.069

^aNumbering of a-b-c in weighted averaging structure represents the weights of SVR, FFNN, and ANFIS models. The numbering of a-b-c in the structure of neural averaging denotes the number of the input layer, hidden layer, and output layer neurons.

^bSince all data are normalized, the RMSE has no dimension.

View Large

This Feature Is Available To Subscribers Only