Neuro-Fuzzy Control for Generalised Nonlinear Systems under a Recursive Framework

This paper proposes a general recursive state- space Neuro-Fuzzy control framework. It combines a eight- layered neuro-fuzzy architecture with a state feedback quadratic stabilising controller. Both the model and controller are updated online within a constrained unscented Kalman filter. Results from a benchmark Multi-Input and Multi-Output system demonstrate the effectiveness of the proposed approach.