Implementation of Adaptive Neuro-fuzzy Model to Optimize Operational Process of Multiconfiguration Gas-Turbines

Table 1

Training error of different MFs.

Membership function (MF)	RMSE output MF type
	Constant	Linear
Product sigmoidal (psigmf)	0.08071	0.004006
Difference sigmoidal (dsigmf)	0.152021	0.0050048
Pi curve (pimf)	0.122082	0.0091116
Gaussian (gaussmf)	0.011867	0.00017438
Gaussian 2 (gauss2mf)	0.050265	0.0022592
Generalized bell (gbellmf)	0.076138	0.0033531
Trapezoidal (trapmf)	0.258162	0.0098235
Triangular (trimf)	0.08071	0.004005