基于希尔伯特黄变换的场地液化识别方法研究  

Research on Site Liquefaction Identification Method Based on Hilbert-Huang Transform

作　　者：杨彦鑫 路华 林子雲 占旭栋 YANG Yanxin;LU Hua;LIN Ziyun;ZHAN Xudong(School of Architecture and Transportation Engineering,Guilin University of Electronic Technology,Guilin 541004,China)

机构地区：[1]桂林电子科技大学建筑与交通工程学院,广西桂林541004

出　　处：《灾害学》2024年第3期36-42,共7页Journal of Catastrophology

基　　金：广西自然科学基金项目“砂土液化场地的侧向扩展计算理论和变形特性研究”(2022GXNSFBA035569);广西研究生教育创新计划项目“地震作用场地砂土液化时间预测研究”(YCSW2023297)。

摘　　要：以场地的地震动记录为分析对象,该文提出了一种基于希尔伯特黄变换(Hilbert-Huang Transform,HHT)的场地液化识别方法。通过地震动记录数据库,结合液化场地频率快速变化和低频占比丰富两个特征,定义了平均卓越频率下降率和低频占比两个参数,并确定场地液化的阈值。将该方法应用于2011年新西兰Mw 6.0基督城地震事件,并与其他三种基于地震动记录的液化识别方法对比。研究结果表明:该文提出的方法在液化识别率为94.4%,高于其他三种方法。为震后液化区域的快速识别提供了新思路。Taking the seismic motion records of sites as the research subject,a liquefaction identification based on Hilbert-Huang Transform(HHT)is proposed.By utilizing seismic record databases and based on the rapid frequency changes and abundant low-frequency components characteristic of liquefied sites,two parameters as mean excellent frequency drop rate and low-frequency ratio are defined and the thresholds for liquefaction of the sites are determined.The method is evaluated by using data from the 2011 Christchurch M6.0 earthquake in New Zealand and is compared with three existing seismic-based liquefaction identification methods.The results indicate that the proposed method could achieve a recognition rate of 94.4%,demonstrating greater performance.For a rapid post-earthquake assessment for liquefied site,this method provides a new idea for liquefaction identification.

关 键 词：液化识别 时频分析 HHT 平均卓越频率下降率 低频占比 

分 类 号：X43[环境科学与工程—灾害防治] X915.5