连续非线性系统的神经网络α阶逆系统控制方法
Neural Network αTh-Order Inverse-System Control Method for Nonlinear Continuous Systems
-
摘要: 对连续非线性系统,采用连续系统的设计方法,利用神经网络对非线性函数的逼近 能力加上用来表征系统动态特性的若干积分器,构造工程上可实现的原系统的神经网络α阶 逆系统(将α阶逆系统直接与原系统串接,构成α阶积分伪线性复合系统),并对α阶积分伪 线性复合系统设计线性控制器,实现对原非线性系统的有效控制.仿真结果表明,该方法适用 于较一般的线性、非线性连续系统,且结构简单,易于工程实现.Abstract: Based on the design process of continuous systems, this paper focuses on the NN ath-order inverse-system of nonlinear continuous system. The NN ath-order inverse-system is constructed by combining the NN approximating nonlinear function of the system with integrators denoting dynamic characteristics. Furthermore, a linear closed-loop controller for the composite system is designed to complete effective control of the original system. Simulation results show that this method can apply to general linear or nonlinear continuous systems ,and that the controller can be implemented easily.
-
Key words:
- Nonlinear system /
- inverse-system /
- neural networks(NN) /
- identification /
- control
计量
- 文章访问数: 2220
- HTML全文浏览量: 170
- PDF下载量: 1469
- 被引次数: 0