Nonlinear systems modeling based on self-adaptive dual control

An enhanced modeling method is proposed for a class of discrete-time nonlinear stochastic dynamical systems for which an explicit linear parameterization of the uncertainty in the dynamics is either unknown or impossible. The architecture employs a network of Gaussian radial basis functions to compensate for the plant nonlinearities. A new method, combining subtractive clustering with GK fuzzy clustering, is adopted to obtain the optimal number and centers of the RBF network selfadaptively. A stable weight adjustment mechanism is then determined using Kalman filtering. Simulation results reveal that the hybrid clustering, used for RBF network training, achieves better performance for nonlinear dynamic systems modeling and function approximation.