摘要
针对仿射非线性大系统 ,采用分散控制的思想 ,利用模糊神经网络辨识方法对各个子系统建立 NARMA模型 ,运用自适应控制、滑模控制方法消除各子系统间的非线性关联项及不确定的外部干扰 ,并使用神经网络的辨识结果来调节控制器参数 ,形成间接自适应分散神经网络控制器 .最后 ,将本文提出的控制方法运用于互连双倒立摆仿真中 。
An indirect adaptive decentralized neurcontrol is proposed for a kind of nonlinear large scale systems by adopting a decentralized control mode. Fuzzy neural network is used to identify the subsystem for building NARMA model and adjusting the parameter of the controller. Besides, the adaptive control and sliding model control method is used to extinguish nonlinear interconnection between each subsystems and disturbances with unknown bounds. The indirect adaptive decentralized neurcontrol controller is verified by its application to a double inverted pendulum system, the results of simulation confirm the effectiveness of the proposed algorithm.
出处
《应用科学学报》
CAS
CSCD
2002年第1期66-69,共4页
Journal of Applied Sciences
基金
国家自然科学基金 ( 699740 2 1)
航空科学重点基金 ( 98z5 10 0 2 )资助项目
关键词
非线性大系统
分散控制
神经网络
自适应控制
nonlinear large scale system
decentralized control
neural network
adaptive control
