Development of Methodology for System Identification of Non-linear System Using Radial Basis Function Neural Network

In the past few years, there has been significant growth in the development and use of artificial neural networks (ANNs). At present ANN technologies are used in such areas of science as pattern recognition, medicine, speech recognition and synthesis, image processing, robotics and control systems. The use of ANN technologies in the control systems refers to develop a complex structure that can be used as intelligence control system and the identification of controlled object. To design intelligence control system and identification models based on ANN, it is necessary to determine the structure and parameters of ANN. One of the promising approaches to solving this problem is the use of evolutionary modeling methods, namely genetic algorithms, for training and structural optimization of ANN. In this paper a synthesis of a Radial Basis Function (RBF) neural network using genetic algorithm which is uses to determine the number of neurons in hidden layer and the parameters of RBF neural network for system identification process of non-linear system was presented.