摘要
该文根据磁盘读取系统的结构,建立了它的物理模型,并指出其存在非线性因素的原因。由于经典控制方法对于非线性系统效果不是很好,因此使用MATLAB神经网络工具箱中的NARMA-L2控制模块进行控制,并利用Simulink可视化建模工具平台设计了整个控制系统。对神经网络控制器进行了训练,并利用经过训练的神经网络控制磁盘读取系统。仿真的结果表明,神经网络NARMA-L2控制模块能够满足含有非线性因素的磁盘读取系统的控制要求。
In this article, the physical model of the HD reading system was built according to its structure,the reasons of why nonlinear factors exist were also explained. Since the classical control method was not suited for the nonlinear system, the NARMA - L2 Model in Neural Network Toolbox of MATLAB was used to control the system. We designed the whole control system by using Simulink,trained the Neural Network controller,used the trained Neural Network to control the HD reading system. The simulation result showed that the NARMA - L2 Model could meet the control requirement of the system having nonlinear factors.
出处
《计算机仿真》
CSCD
2005年第8期202-205,共4页
Computer Simulation
关键词
磁盘读取系统
非线性
神经网络控制
仿真
HD reading system
Non - linearity
Neural - network control
Simulation
