Nonlinear system identification based on delta-learning rules

The neural network can be used to identify unknown systems. A novel method based on delta-learning rules to identify the nonlinear system is proposed. First, a single-input-single-output (SISO) discrete-time nonlinear system is introduced, and Gaussian basis functions are used to represent the nonlinear functions of this system. Then the adjustable parameters of Gaussian basis functions are optimized by using delta-learning rules. In the end, simulation results are illustrated to demonstrate the effectiveness of the proposed method.