摘要
如何利用近断层强震记录有效获取速度、位移时程,恢复永久位移信息是研究近场强地震动特性需要首先解决的问题。由于背景噪声、强震引起的仪器倾斜等因素,近断层强震记录中存在严重的基线漂移,迫切需要发展直观、有效的基线校正方法。在由Iwan方法发展而来的V0校正算法基础上,引入了"时移斜率比"、"拟合段标准差"、"位移尾段均方差"等参数,提出改进的基线校正优化方法及其具体操作步骤。利用汶川地震近断层强震记录对方法进行检验,获得的永久位移信息与GPS测量结果吻合较好。与已有V0校正算法相比,文中方法能直观有效地获得速度时程尾段直线拟合的最佳时间范围,一定程度上减少人为确定该时段给校正结果带来的不确定性。
Deriving velocity and displacement time histories from near-fault observation s t rong motion records and recovering the permanent displacement information is the first obstacle need to be solved while studying the near-field strong ground motion characteristics. Most near-fault observation strong motion records are plagued by baseline drifts aroused by background noise or tilt of instruments due to strong motion triggering. Thus, there is an urgent need to develop an intuitive and effective baseline correction method. Iwan proposed an adjustment procedure for removing baseline drifts from accelerograms in 1985, the further modification to Iwan et al.'s original method is termed by Boore(2001), the v0 correction method. Our discussion is based on v0 correction method. At first, we introduced consecutive slope ratios, standard deviations of fitted velocity segment and mean square error of adjusted displacement segment. With the aid of these three parameters, we proposed an improved baseline correction optimization method and specific steps. Observation strong motion records of Wenchuan Ms8.0 earthquake are used to verify this method. Permanent displacement information obtained from records are in good agreement with GPS measurements. Compared with v0 baseline correction method, proposed optimized method can determine fitted range of final segment of velocity with linear trend. To some extent, it could reduce baseline correction uncertainty which comes from determining fitted range subjectively.
出处
《土木工程学报》
EI
CSCD
北大核心
2014年第S2期300-306,共7页
China Civil Engineering Journal
基金
国家自然科学基金重大研究计划集成项目(91215301)
国家自然科学基金青年项目(51208474)
地震科技星火计划项目(XH13039Y)
中央公益型院所基本业务专项项目(ZDJ2014-07)
关键词
近断层
强震观测记录
基线校正
优化
near-fault
observation strong motion records
baseline correction
optimization