基于神经网络的磁轴承自适应控制器  

Active Magnetic Bearing Adaptive Controller Based on Neural Network

作　　者：徐春广[1] 武涵[1] 吕冬明[1] 肖定国[1] 郝娟[1] 

机构地区：[1]北京理工大学机械与车辆学院,北京100081

出　　处：《机床与液压》2009年第6期135-137,共3页Machine Tool & Hydraulics

摘　　要：研究机床电主轴上采用的主动磁轴承系统的建模,设计了基于RBF神经网络的自适应PID控制器。该控制器可根据磁轴承的运行情况进行在线自学习和自校正,避免了PID控制器参数整定困难且无法及时改变的缺点。基于某磁轴承系统的仿真和实验研究表明,该RBF-PID控制器能很好地抑制转子振动,满足控制要求。The modeling and controller design of an active magnetic bearing （AMB） system were studied, which is used in motorized spindle. Based on radial basis function （RBF） neural network, an adaptive PID controller was designed. According to run- ning state of active magnetic bearing, the controller is self-learning and self-turning. Simulation and experimental results show that RBF-PID controller has better performance on inhibiting vibration of AMB rotor than PID controller. The control requirements of the system are satisfied.

关 键 词：主动磁轴承 神经网络 自适应控制 

分 类 号：TP273.2[自动化与计算机技术—检测技术与自动化装置]