神经网络异步自学习控制系统
Neural Network Based Asynchronous Learning Control Systems
-
摘要: 基于异步自学习控制方法,利用前馈网络对学习动态特性建模,从而将两者结合起来,既 避免了前者对重复性的苛求,又避免了神经网络控制方法通常存在的分析与实时控制的困难. 文中证明了整个系统的稳定性,并以机械手为例进行了仿真.
-
关键词:
- 神经网络与智能控制 /
- 异步自学习控制 /
- Lyapunov稳定性理论 /
- 学习算法 /
- 机器人控制
Abstract: In this paper the neural network based asynchronous learning control system, on the basis of the asynchronous learning control method given by ref. [1], is proposed. The gradient-type learning control algorithm is derived, the strict proof on stability convergency is provided by Lyapunov stability theory, and the simulation study of two links of PUMA 560 robot systems is given. The results show that the RMS track accuracy of this proposed method is improved greatly compared to classical PID control. -
计量
- 文章访问数: 2904
- HTML全文浏览量: 68
- PDF下载量: 1104
- 被引次数: 0
下载:
