摘要
为解决传统磁流变液阻尼器(Magneto Rheological Fluid Damper,MRFD)磁场利用率低及磁流变液沉降导致控制特性劣化难题,提出新型阻尼器—磁流变脂阻尼器(Magneto Rheological Grease Damper,MRGD)。采用神经网络(Neural Network,NN)对足尺MRGD动力特性进行辨识,通过将改进的限幅最优(Modified Clipped-Optimal,MCO)算法整合到模糊神经网络(Fuzzy Neral Network,FNN)理论来设计适合MRGD的FNN/MCO半主动控制策略,并构建SIMULINK仿真分析平台。以典型三跨铁路连续梁桥为工程背景,分别对未控制、FNN/MCO半主动控制及线性二次型高斯(Linear Quadratic Gaussian,LQG)主动控制下桥梁各项评价指标进行分析。结果表明,所提FNN/MCO半主动控制策略对桥梁地震响应控制效果明显优于LQG主动控制策略;FNN/MCO策略较LQG策略更利于控制装置性能发挥;FNN/MCO策略稳定性、鲁棒性均明显优于LQG策略。
In view of the problems of low utilization rate of magnetic field and degraded control characteristic due to sedimentation of magneto rheological fluid in traditional magneto rheological fluid damper （MRFD）,magneto rheological grease damper （MRGD）as a new type of damper was proposed.The neural network （NN）was employed to identify the dynamic characteristics of a full-scale MRGD,the FNN /MCO semi-active control strategy for MRGD was designed by integrating the modified clipped-optimal （MCO） algorithm with the fuzzy neral network （FNN ） theory,and the SIMULINK simulation analysis platform corresponding to FNN /MCO strategy was constructed.Taking the typical three-span continuous girder railway bridge as engineering background,the various evaluation criteria for the bridge with non-control,FNN /MCO semi-active control and active control based on linear quadratic Gaussian （LQG）strategy were analyzed respectively.The analytical results show that the control effect on seismic responses of bridge with the FNN /MCO semi-active control strategy proposed is obviously superior to those with LQG active control strategy.Comparing with LQG strategy,the FNN /MCO strategy can more obviously contribute to the performance exertion of control devices.The stability and robustness of the FNN /MCO strategy are both superior to those of LQG control strategy.
出处
《振动与冲击》
EI
CSCD
北大核心
2015年第2期66-73,共8页
Journal of Vibration and Shock
基金
"十二五"国家科技支撑计划重点资助项目(2012BAJ14B00)
国家自然科学基金重点资助项目(51078077)
关键词
铁路桥梁
磁流变脂阻尼器
模糊神经网络
改进的限幅最优控制
半主动控制
railway bridge
magneto rheological grease damper
fuzzy neural network
modified clipped-optimal control algorithm
semi-active control
