时间序列模型中的随机共振现象  

Stochastic Resonance Phenomenon Study in Time Series Model

作　　者：何俊[1] 陈怀海[1] 贺旭东[1] 

机构地区：[1]南京航空航天大学航空宇航学院南京,210016

出　　处：《振动．测试与诊断》2014年第1期108-112,193,共5页Journal of Vibration,Measurement & Diagnosis

基　　金：国家自然科学基金资助项目(111102083);民机专项科研资助项目(MJ-G-2011-24)

摘　　要：在滑动自回归(auto-regressive and moving average,简称ARMA)时间序列模型的基础上,利用模态稳定性图来确定系统真实模态,描述了在求解过程中产生的随机共振现象。借助悬臂梁的有限元模型,利用精细时程积分方法计算得到了其加速度脉冲响应函数,建立了用于振动模态识别的ARMA模型。在利用模态稳定性图来确定系统真实模态的过程中发现,加入合适的噪声信号可以有效地改善识别结果,剔除虚假模态,即产生了随机共振现象。对悬臂梁进行时变化处理后,随机共振现象较之前不变系统更加显著,对最终识别结果产生了明显的优化作用。A detailed description of the stochastic resonance phenomenon is given,which is generated in process of determining the real modal of the system by modal stability diagram method based on auto-regressive and moving average(ARMA)time series model.The precise time-integration method is used to calculate the acceleration pulse response function on the finite element model of the cantilever,and then establish the ARMA model for vibration modal identification.It is found that the recognition results can be effectively improved by adding appropriate noise signal and the false modes can be eliminated in process of determining the real modal of the system by modal stability diagram method,namely the stochastic resonance phenomenon.After time-varying treatment of the cantilever beam,the stochastic resonance phenomenon is more significant compared with the time-invariant system and generate an obvious optimization to the final results.

关 键 词：滑动自回归模型 稳定图 随机共振 时变系统 模态频率 

分 类 号：O328[理学—一般力学与力学基础] TH17[理学—力学]