Nonlinear System Identification Using Lyapunov-Based Fully Tuned RBFN
In recent years, nonlinear system identification using neural network has become a widely studied area because of its close relationship to the system control. Basically, any good identification scheme that incorporates RBFN should satisfy two criteria: (i) The parameters of the RBFN are tuned appropriately so that its output to an input signal can approximate the response of the real system to the same input with good accuracy. (ii) The network structure is compact and the parameter adaptive law is efficient so that fast on-line learning can be implemented.
KeywordsHide Neuron Extend Kalman Filter Radial Basis Function Neural Network Dead Zone Hide Unit
Unable to display preview. Download preview PDF.