Mathematical Model Identification of Robot Drive System Based on Neural Network

This paper studies mainly a wheeled mobile robot drive system with modeling method proposed based on combination of neural network analysis and experiment.Through the modeling method which is based on analysis of the ideal DC motor mathematic mode,the mathematical model identification of robot drive system has been achieved by using neural network with serial-parallel identification to confirm the relationship between the input and output of the system.This method has been proved available through experiment,and will be the technical support of the future research of accurate motion and drive control in the industrial and agricultural field,etc.