تشخیص زمینلرزه و انفجار به کمک تبدیل موجک نمای هورست محلی

In this paper, the application of fractal methods of local Hurst exponent and their wavelet transform, to differentiation between time series seismograms of earthquake and explosion is considered. By using the Hurst exponent, short or long-range memory in a time series, would be determined and for their local time behavior, the local Hurst exponent is applied instead of typical Hurst exponent. In this case Daubechies' wavelets with fractal structure can be helpful for better analysis. In this study, first we consider three important factors: discreteness, long-range memory and local- time behavior, for seismic signals recorded on the seismograms, then the Hurst exponent in typical time feature, has been transferred to time-frequency environment for doing better analysis. This approach, permitting us to obtain an algorithm for noisy and much damping data, in comparing with the normal time dependent method.