-
摘要: 针对非线性离散时间系统的控制问题,提出了一种基于近似模型的多层模糊CMAC 自适应控制方法.采用多层模糊CMAC对非线性函数进行逼近,并提出了一种新的神经网络学 习算法来保证权值的有界性.由于无需满足PE条件,所以文中提出的方法对于离散时间系统 的神经网络控制问题具有实际价值.Abstract: A multi-layer fuzzy CMAC adaptive control method based on an approximate model is presented in this paper for nonlinear discrete-time systems. The nonlin- ear functions are approximated by multi-layer fuzzy CMAC. A new training algorithm is proposed to guarantee the bounded weight, and persistency of excitation condition is not required. The presented algorithm has practical value for neural network control of discrete-time system.
-
Key words:
- Neural networks /
- nonlinear control /
- approximation model /
- adaptive control
-
计量
- 文章访问数: 2912
- HTML全文浏览量: 128
- PDF下载量: 963
- 被引次数: 0