Fuzzy neural network with general parameter adaptation for modeling of nonlinear time-series

By taking advantage of fuzzy systems and neural networks, a fuzzy-neural ne twork with a general parameter (GP) learning algorithm and heuristic model structure determination is proposed in this paper. Our network model is bas ed on the Gaussian radial basis function network (RBFN), We use the flexibl e GP approach both for initializing the off-line training algorithm and fin e-tuning the nonlinear model efficiently in on-line operation. A modificati on of the robust Unbiasedness Criterion using Distorter (UCD) is utilized f or selecting the structural parameters of this adaptive model. The UCD appr oach provides the desired modeling accuracy and avoids the risk of over-fit ting, In order to illustrate the operation of the proposed modeling scheme, it is experimentally applied to a fault detection application.