主动磁悬浮轴承的神经网络自适应稳定性控制

针对主动磁悬浮轴承本质非线性和开环不稳定的系统特征,设计了一种BP神经网络自适应PID控制器。该控制器采用改进的BP神经网络PID控制算法,通过BP神经网络的自学习和权值调整寻找最优的PID参数,克服了常规PID控制参数整定困难的缺陷,实现了系统的自适应控制。通过MATLAB/Simulink环境和S-Function模块建立了主动磁悬浮轴承控制系统模型,并进行了系统仿真实验,结果表明,BP神经网络自适应PID控制系统响应速度更快,具有更好的动态性能和稳态性能。

Adaptive NeuraI Network StabiIity ControI of Active Magnetic Bearing

Aiming at the intrinsic nonIinearity and open Ioop instabiIity of the active magnetic bearing system,an adaptive PID controI er based on BP neuraI network is designed in this paper.The controI er adopts improved PID controI aIgorithm to find the optimaI PID parameters by Iearning the BP neuraI network and adjusting weights.The controI er aIso overcomes the parameters tuning difficuIty of conventionaI PID controI er,so as to reaIize the adaptive controI.The controI system modeI of the active magnetic bearing is estabIished by the MATLAB / SimuIink environment and S-Function moduIe.Moreover,the sys-tem simuIation experiments are aIso carried out.