基于异常检测与结构性态的加速度脉冲识别方法

ANOMALY DETECTION AND STRUCTURAL PERFORMANCE-BASED IDENTIFICATION METHOD FOR ACCELERATION PULSE

  • 摘要: 从工程地震学角度,近断层强震记录中所包含的加速度脉冲是造成结构破坏的重要原因,识别这类地震动的主要目的是服务于工程结构的抗震设防。该文提出了借助时间序列异常检测技术初步提取加速度脉冲波形,以其作用下结构性态为加速度脉冲型地震动的最终识别标准。在NGA-West 2数据库中筛选矩震级大于5.5,断层距小于60 km的2390条强震记录,通过静力弹塑性分析方法求解Hazus中35类典型结构响应并据此识别出加速度脉冲型地震动。结果表明:加速度脉冲型地震动对自振周期1 s内的结构影响大。时频分析表明:加速度脉冲型地震动的中心频率主要分布在1 Hz以上。该文提出的识别方法能够有效识别原始地震动中不规则的脉冲波形,以结构性态作为脉冲型地震动判别标准,具有更好的工程适用性。

     

    Abstract: From the view of engineering seismology, the near-fault records with acceleration pulses are the critical causation to structural damage. Thusly, the identification of pulse-like records is essential for seismic design. Anomaly detection is applied to extract pulse waveform preliminarily in the detection, and structural performance serves as the criterion to identify acceleration pulse-like ground motion. A total of 2390 records with rupture distance less than 60 km and moment magnitude exceed 5.5 are collected from the NGA-West2 database, and the structural performance of 35 modeling structures in Hazus is analyzed by the pushover analysis method. Acceleration pulse-like ground motion has a significant amplification effect on structures with fundamental period Te less than 1 s. The central frequency of acceleration pulse-like motions mainly concentrates above 1 Hz. The method proposed can effectively identify the irregular pulse waveform in the original ground motion and the structural performance-based criterion is more applicable in earthquake engineering.

     

/

返回文章
返回