/Neuro_fuzzy

n this work, a neuro-fuzzy system (NFS)-based autonomous and adaptive controller termed as the hybrid auto-adaptive controller (HAC) is developed. HAC combines a simplified NFS (Simp_NFS) and a simplified neural network (Simp_NN). Unlike conventional NFS, in this study, hyper-plane-shaped clusters (HPSCs) are utilized in Simp_NFS, which has no learning parameters like mean and variance in the antecedent part. Only the consequent parameter needs to be adapted, where the adaptation laws are derived from the Simp_NN. The number of learning parameters in Simp_NN reduced to one by replacing the weights between the hidden and output layers with their mean value.

Primary LanguageMATLABMIT LicenseMIT

Watchers