Real-Time Identification of Nonlinear Systems Using MRAN/EMRAN Algorithm
Minimal resource allocating network (MRAN) is a recent algorithm for implementing fully tuned RBF network. Unlike the derived parameter tuning rules in Chapter 3, in MRAN algorithm, an extended Kalman filter (EKF) is utilized to update all the parameters of the RBFN. Although lacking a strict mathematical proof, MRAN was shown to be more effective than other algorithms (like RAN and RANEKF) in function approximation and pattern classification .
KeywordsHide Neuron Extended Kalman Filter Radial Basis Function Neural Network Hide Unit Flight Control
Unable to display preview. Download preview PDF.