论文部分内容阅读
对四种型号的数字强震仪（ＳＳＲ１／ＦＢＡ１３，ＳＳＲ１／ＦＢＡ２３，Ｋｅｌｕｎｊｉ／ＳｌＪ１００和ＳＣＱ１）作了振动台试验分析。发现加速度记录的背景噪声水平均较低，但均存在零线变动现象。所产生的长周期误差对加速度记录影响极小，但使积分位移产生很大的零线漂移，因而无法直接积分得到可靠的位移时程，在对不同校正方法进行了比较研究之后，提出了一种校正处理方法。首先对加速度记录按Ｉｗａｎ等人的方法作零线校正，然后积分速度时程，在积分位移前对速度零线进行调正，最后对积分位移用Ｂｕｔｔｅｒｗｏｒｔｈ滤波器作一次高通滤波，截止频率根据信号与噪声的傅氏谱的比较确定。校正结果与实际位移时程符合很好。
Shaking table tests were conducted on four models of digital seismographs (SSR1 / FBA13, SSR1 / FBA23, Kelunji / SlJ100 and SCQ1). It is found that the background noise level of acceleration record is low, but there is zero line fluctuation. The long-period error produced has little effect on the acceleration record, but it makes the integral displacement have a great zero-line drift, so it can not be directly integrated to get a reliable displacement time history. After comparing and researching different correction methods, Kind of correction processing method. First of all, Iwan et al. Record acceleration records for zero line correction, and then integrate the velocity time series, before the integral displacement of the zero line speed correction, and finally for the integral displacement Butterworth filter for a high-pass filter, the cutoff frequency based on the signal The comparison with the noise Fourier spectrum is confirmed. Correction results and the actual displacement time is in good agreement.