Adaptive Spread Coefficient-based RBF-NN for Complex Signals Modeling

As an efficient method on the fitting or approximating for complex signals, the Radial Base Function Neural Network (RBF‐NN) is widely used in signal modeling. During the training process, the spread coefficient (Sc) is one of important parameters in the RBF‐NN learning algorithm. A suitable Sc can speed up the signal fitting process. This paper presents an improved RBF‐NN learning method based on the adaptive spread coefficient for the signal approximation of complex systems. The improved algorithm is applied to the learning and approximating process of the nonlinear signal. The simulations showed that the presented RBF‐NN has good effects on speeding up the training and approaching process. Meanwhile, the learning convergence of the improved algorithm is more excellent than that of normal algorithm.