摘要
基于频率变化的结构地震损伤评估方法具有机理明确和精度较高等特点,但传统信号分析方法在时频分辨率上不能同时满足精度要求,导致时变频率不能直接从响应信号中精确获取,影响频率法损伤评估的应用。依据时频边缘条件提出时频谱分析精度评价标准,通过对比不同的信号分析方法,确认具有特定基函数的小波包变换是获取精确时变功率谱的有效工具;提出基于小波包脊的时变频率提取方法,在此基础上依据结构频率的变化可计算结构时变损伤指标,并最终实现结构多维地震损伤评估。算例表明基于小波包变换和时变频率的结构地震损伤评估方法可以较准确地反映结构的整体损伤演变过程和最终损伤程度;应用该方法时仅需结构的位移时程,在结构动力分析、抗震验算及实际结构的震害评估中均具良好的适用性。
The structural seismic damage assessment method based on frequency changes has the advantages,such as,clear mechanism and high precision,but the time-frequency resolution for the traditional methods of signal analysis can not accurately meet the requirements of precision. Hence,the time-varying frequencies cannot be obtained directly from the response signals and the application of this damage assessment method is affected. Here,accuracy evaluation criteria for different signal analysis methods were presented based on marginal conditions in time-frequency domain,and the wavelet packet decomposition method with special wavelet basis functions was verified as an efficient tool to establish time-varying power spectrum on the basis of theoretical and computational analyses. The time-varying frequency extraction method based on wavelet packet ridge was proposed,and the structural time-varying seismic damage index was calculated according to structural frequency changes,furthermore,the structural multi-dimensional seismic damage assessment was realized. The examples showed that the seismic damage assessment method based on wavelet packet transformation and time-varying frequencies can accurately reveal the overall damage evolution process and the eventual damage level of a structure; this method only requires structural displacement histories,and has a good applicability for structural dynamic analysis,seismic design verification and seismic damage assessment of actual structures.
出处
《振动与冲击》
EI
CSCD
北大核心
2016年第7期23-30,共8页
Journal of Vibration and Shock
基金
国家自然科学基金项目(51478024)
工程抗震与结构诊治北京市重点实验室重点项目(USDE201403)
关键词
地震
损伤评估
小波包变换
时变频率
时变功率谱
earthquake
damage assessment
wavelet packet transformation
time-varying frequency
time-varying power spectrum