基于声波与振动探测的地震灾害生命搜索系统信号分析
|
摘要
基于声波与振动探测原理的地震生命探测系统的研制在我国灾害生命搜索救助领域中占有重要位置。对于其中数理模型的研究是研制和开发声波、振动生命探测定位仪的基础。测试实验表明,在声波与振动灾害救助系统中,所采集的信号带有大量的随机噪声,为了提高信噪比,必然要进行滤波。本文用的是小波变换的去噪方法。同时,从传感器接收到的信号可能是从多个振源发出的,这就要涉及到多源信号的分离问题。为此,引入了独立变量分析(ICA)的分离方法和傅立叶变换加以解决。当找到需要的信号后,通过相关函数分析自动获取多道信号的相对时差,进而求取振源的位置。
The research on the victim location system based on acoustic and seismic wave detection plays an important part in post-earthquake search and rescue. This project will contribute to the life search and rescue field in China, especially on the earthquake emergent life salvation. In the disaster rescue system based on acoustic/seismic waves’ detection, the signals from the victim’s knocking underground are often combined with a great number of random noises. In order to increase the SNR (Signal Noise Ratio), the signal filter by using the Wavelet Transform (WT) method has been accomplished. Meanwhile, the signal that we received from the sensor may be the mixed signals coming from several signal sources. For this reason, we have to consider about how to separate the kind of mixed signals. So, we introduce the Independent Component Analysis (ICA) and Fourier Transform (FT) method to solve this problem. After dissociating the effective signals, we can locate the vibration source according to the time difference that can be got by using the analysis of correlation-function.
The research on the victim location system based on acoustic and seismic wave detection plays an important part in post-earthquake search and rescue. This project will contribute to the life search and rescue field in China, especially on the earthquake emergent life salvation. In the disaster rescue system based on acoustic/seismic waves’ detection, the signals from the victim’s knocking underground are often combined with a great number of random noises. In order to increase the SNR (Signal Noise Ratio), the signal filter by using the Wavelet Transform (WT) method has been accomplished. Meanwhile, the signal that we received from the sensor may be the mixed signals coming from several signal sources. For this reason, we have to consider about how to separate the kind of mixed signals. So, we introduce the Independent Component Analysis (ICA) and Fourier Transform (FT) method to solve this problem. After dissociating the effective signals, we can locate the vibration source according to the time difference that can be got by using the analysis of correlation-function.
引文
[1]Hyvarinen A,Oja E.Afast fixed-point algorithm for independent component analysis[J].Neural Com-putation,1997,(9):1483-1492.
[2]周卫东,贾磊,李英远.一种独立分量分析的迭代算法和实验结果[J].生物物理学报,2002,18(1):57-60.
[3]杨竹青,胡德文.独立成分分析方法在盲源信号分离中的应用[J].计算机测量与控制,2002,10(3):200-202.
[4]李力,屈梁生.应用独立分量分析提取机器的状态特征[J].西安交通大学学报,2003,37(1):45-48.
[5]S Shamsunder,G B Giannakis.Modeling of non-gaussian array data using cumulants:DOAesti mation of more sources withless sensors[J],Signal Process-ing,1993,30(3):279-297.
[6]I Daubechies.Ten lectures on wavelets[M].CBMS,SIAM,61,1994.194-202.
[7]Jeanny Herault,Christian Jutten.Space or ti me a-daptive signal processing by neural network models[M].Neural Networks for Computing,Snowbird Utah,1986.
[8]王绪本,郭勇,曹辉,等.城市灾害生命救助中的声波与振动数理模型试验研究[A].中国地球物理学会年刊[C].北京:地震出版社,2002.
[2]周卫东,贾磊,李英远.一种独立分量分析的迭代算法和实验结果[J].生物物理学报,2002,18(1):57-60.
[3]杨竹青,胡德文.独立成分分析方法在盲源信号分离中的应用[J].计算机测量与控制,2002,10(3):200-202.
[4]李力,屈梁生.应用独立分量分析提取机器的状态特征[J].西安交通大学学报,2003,37(1):45-48.
[5]S Shamsunder,G B Giannakis.Modeling of non-gaussian array data using cumulants:DOAesti mation of more sources withless sensors[J],Signal Process-ing,1993,30(3):279-297.
[6]I Daubechies.Ten lectures on wavelets[M].CBMS,SIAM,61,1994.194-202.
[7]Jeanny Herault,Christian Jutten.Space or ti me a-daptive signal processing by neural network models[M].Neural Networks for Computing,Snowbird Utah,1986.
[8]王绪本,郭勇,曹辉,等.城市灾害生命救助中的声波与振动数理模型试验研究[A].中国地球物理学会年刊[C].北京:地震出版社,2002.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心