Research On Phase Automatic Identification Of Earthquake Early Warning

Earthquake disaster is the chief culprit of natural disasters, the disasters causedby great earthquake have the characteristics of sudden and destructive, and willproduce serious secondary disasters and bring severe damage to people’s lives andproperties. In order to mitigate effectively the casualties and property losses caused byearthquake, now besides strengthening earthquake monitoring and prediction andengineering structure seismic design, earthquake early warning is a new and effectivemeans of protecting against and mitigating earthquake disasters, which has importantsignificance in mitigating earthquake losses, reducing the secondary disasters and thecasualties. In this paper, we study on seismic phase automatic identificationtechnology of earthquake early warning, and make the following aspects of researchprogress and results:1、Aiming at defects on the application of the traditional time-frequency forseismic phase identification, we introduce new time-frequency method, theHilbert-Huang transform, will be applied to the seismic data signal processing. Firstlythe paper studies in detail the principle of Hilbert-Huang transform and the existingproblems about HHT; then HHT is used to analyze several typical non-stationarysignals in the time-frequency domain, the results show that the HHT method candepict well local time-frequency characteristics of signals, has the ability of detectingbreak points and high time-frequency resolution; Finally, we apply HHTtime-frequency analysis for seismic signal with non-stationary dynamic changecharacteristic, it can be seen that HHT method can sensitively extract to the maincharacteristics of seismic wave signal dynamic change over time and frequency, thetime-frequency spectrum is clear, and can reflect the information with the energy’sabrupt change points, the time-frequency local orientation ability is strong.2、HHT method is applied to calculate the instantaneous frequency of seismicsignal, based on the instantaneous amplitude and instantaneous frequency, we buildthe new STA/LTA characteristic functions on identification of P wave and S wave.The results show that the new characteristic functions can fully embody thecharacteristics that amplitude increases and frequency changes when seismic first-break waves and subsequent phases arrive.3、Based on the research, seismic phase automatic identification method at homeand abroad of earthquake early warning, in this paper, applying new STA/LTAcharacteristic functions, we put forward multi-step phase identification method basedon STA/LTA method and AIC method to P-phase and S-phase identification. To solvethreshold selection problem of STA/LTA method, this paper puts forward to takedirectly the moment the maximum STA/LTA ratio arrives, which also is the roughtime P wave and S wave arrive, the seismic phase automatic identification resultsshow that this processing technology is effective. Application of local seismic datafrom broadband mobile array in southeast region of Gansu Province, verifying themethod in this paper, compared with artificial recognition results and the results ofother seismic phase automatic identification methods, the proposed approach of Pwave and S wave automatic identification can obtain higher recognition accuracy, andno matter P-phase and S-phase automatic identification, multi-step phaseidentification method based on STA/LTA and VAR-AIC is superior to multi-stepphase identification method based on STA/LTA and AR-AIC.