【作者】 于阿涛；

【导师】 赵鸣；

【作者基本信息】 同济大学 ， 结构工程， 2007， 硕士

【摘要】 我国处于环太平洋地震构造系和大陆地震构造系的交汇部位，因此我国的地震频繁而强烈。地震除了造成工程破坏等直接损失之外，还会引起如火灾等次生灾害，造成巨大的经济损失。为了减少地震造成的损失，本文研究了振动特征传感器，该传感器可用于重大工程和生命线系统工程，该传感器每隔一段时间计算一次场地振动特征，根据计算的场地振动特征进行地震报警，防止由于仅依靠振动幅值进行地震报警而引起的误判。振动特征传感器计算的振动特征还可以指导地震后救灾活动。本文工作主要围绕特征传感器进行，重点放在特征传感器各个功能模块的选取上：信号预处理、振动特征的选取。此外，研究了根据场地地震加速度记录识别场地液化的方法。信号预处理部分：对于压电传感器测量的加速度信号，在分析信号之前，需要对信号进行预处理，信号预处理针对信号中的趋势项进行。分析几种常用的消除趋势项的滤波方法，得出结论：可以采用Butterworth高通滤波器和最小二乘法滤波对信号趋势项进行预处理。振动特征部分：分析了261条地震竖向记录，采用基于统计方法识别P波初动，并对P波初动后3秒内的振动特征进行分析，提出场地速度、谱成分、过零周期等振动特征区分地震与非地震振动；分析实测的489条非地震振动和1642条地震水平向记录，分析整个振动时程，提出谱成分、过零周期等振动特征区分地震与非地震振动。场地液化识别：参考现有的根据场地加速度记录分析场地液化的方法，采用小波分解和Hilbert谱分析求解场地的瞬时频率，根据场地瞬时频率识别场地的液化情况。最后，根据以上的研究，理论上提出了分析全时程振动的特征传感器结构示意图，以及特征传感器各部分的功能。更多还原

【Abstract】 China locates in the joint of the Pacific earthquake zone and the continentalearthquake zone, so the earthquakes in our country are of high frequency and of highintensity, The earthquakes not only result in the structural destroys but also induce thesecond disasters, such as fire and so on, which inducing much lose. The vibrationcharacter sensor using the ground acceleration to quickly identify the earthquake isstudied in this paper. The vibration character sensor can be used by the importantengineering and the lifeline engineering. The vibration character sensor calculates thevibration characters though which the earthquake is recognized. The vibrationcharacter sensor can reduce the chance making false alarm and offer the earthquakerescue work guidance.The key work of this paper is studying the vibration character sensor andchoosing the calculation methods, such as the processing of the crude signal, and thevibration character. Then the methods identifying soil liquefaction through theearthquake acceleration time history record are studied.The processing of crude signals: the crude signals must be processed to removethe baseline drifts and to be filtered before used. Several methods are discussed, thehigh passing Butterworth filter can be used for real time signal processing, and theleast square error method can be used when real time signal processing is notrequired.Vibration character: 261 earthquake vertical records are used to identify the Pphase of the earthquake using the statistical method. And the 3 seconds records afterthe P phase arrival are analyzed to find some vibration characters, such as spectrum,ground velocity, zero crossing time. 489 records other than earthquake records and1642 earthquake horizontal records are analyzed to fmd some vibration characterssuch as spectrum and zero crossing time to identifying the earthquake from others.Soil liquefaction identifying: several methods identifying the soil liquefactionare discussed. This article chooses the wavelet transform and the Hilbert spectrumtransform to get the ground instantaneous frequency through which the soilliquefaction is identifying.In the end, the sketch map of the vibration character sensor which analyzes thewhole record and the functions of all the parts are theoretically constructed.更多还原