Identification of nonlinear dynamic systems using neural networks

In this paper it presents a specific introduction to basic knowledge and development condition of the neural network. It also raises that in allusion to the non-linear and slow time varying characteristics of D.C. motor model parameter, it may utilize BP network to set up rotating speed identification model of D.C. motor, its network study adopts error feedback study algorithm. This kind of network has strong ability of expressing and processing instantaneous information, applicable for solving the problem of non-linear dynamic identification and control. The simulation result of adopting MATLAB software package to the system shows that this rotating speed identification model has good response capability and it can follow the changing track of the motor's rotating speed accurately.