摘要
系统中分别设计了一个神经网络系统辨识器(NNP I)和一个模型参考自适应神经网络控制器(NNP IC),NNP I自适应地在线辨识出系统的集中不确定量,NNP IC能使到系统输出跟踪参考模型的输出。仿真实验表明,与常规的控制器相比,本文设计的速度控制方案能取得优良的控制性能,且在负载转矩和电机内部参数变化的情况下有很强的鲁棒性。
The Neural-Network Plant Identifier (NNPI) and the Model Reference adaptive Neural- Network controller(NNPIC) are designed in the proposed speed control scheme respectively. NNPI is used to identify the lumped uncertainty and NNPIC is used to bring the output of the system to the output of the reference model. Computer simulations demonstrate that the proposed scheme can obtain satisfied performances and a robust control compared with the conventional controller.
出处
《佛山科学技术学院学报(自然科学版)》
CAS
2005年第4期25-28,共4页
Journal of Foshan University(Natural Science Edition)
基金
广东省自然科学基金资助项目(020118)
关键词
矢量控制
感应电动机
神经网络
自适应控制
field-oriented control
induction motor
neural-network
adaptive control