• 2018-07
  • 2018-10
  • 2018-11
  • Introduction A robust estimate of DOA


    Introduction A robust estimate of DOA of a seismic source is usually accomplished using an array of seismic geophones at a relatively large distance from the source [1–10]. Bearing information can also be obtained using a single geophone with three orthogonal sensing elements that provide polarization-discriminating information related to the direction of propagation of the seismic wave—S-wave and P-wave travel at different velocities, giving rise to time delays—which can be converted to input information for DOA computation [11–14]. However, at the short distance, the array configuration is not effective because of the near-field nature of the seismic wave. On the other hand, a single hiv protease inhibitors capable to sense the different polarizations of the wave can still work, provided that the algorithm does not rely on time-delay information (which might be too small to detect at such short distance). A new detection scheme has to be devised, along with its own algorithm that satisfies the usual requirements for DOA application. Among these requirements is that the DOA estimation algorithm must be computationally efficient in order to speed up processing and reduce power consumption. Based on our previous work on detection of ground vibration and seismic activities using FBG [15–17], we have developed a new approach to tackle the problem of estimate of DOA at short distance with a single tri-axial seismic geophone. This represents an innovative, robust, and simple algorithm for obtaining bearing information on the seismic events, such as people walking or vehicles moving. The proposed method of DOA estimate is based on the interaction and projection of surface-propagating seismic waves generated by the moving personnel or vehicles with a single tri-axial seismic sensor. The distance between the source of the seismic wave and the detector is less than or comparable to one wavelength (less than 100 m) so that the far-field considerations are not valid.
    Tri-axial FBG seismic geophone
    Surface-wave based algorithm for DOA
    DOA experiment of a tri-axial FBG geophone We have carried out some preliminary experimental study to investigate the possibility of using a single tri-axial geophone for obtaining DOA estimate. The geophone used is a fiber optic Bragg grating based three-axis seismic geophone described above. Fig. 8 shows the experiment setup, with a person walking directly to the geophone from 100 m away. Fig. 9 shows the geophone response without any signal processing, where we could only see a few high quality signal cycles. Fig. 10 shows that more walking signal is detected by our processing method of signal correlation even though the signal is buried under noise. Correlation detection is an effective demodulation method for weak signal, which is based on the convolution between unknown and known signals. If parts of the two signals are coherent, the unknown signal can be detected even if it is very weak or buried under noise. Compared with direct detection, the correlation detection has much higher signal-noise-ratio and excellent frequency selectivity. For the application of DOA, due to the walking signal weakness, especially from far away, and the environment\'s complexity, i.e., multi-state noise jamming, the correlation detection is a preferred detection method. Fig. 11 shows that a walking signal is detected by a single tri-axial geophone; Fig. 12 shows the signal after a 50 Hz low-pass filter. It can be seen from Fig. 12 that the signal to noise ratio (SNR) was improved by the low-pass filter, and the difference between North/South (NS)-axis and the East/West (EW)-axis is now apparent, making an estimate of DOA based on the detected signal possible. Fig. 13 describes a second experiment, where the seismic signals are generated by dropping a metal-ball from a given height. To make sure that we get the similar seismic signal, the height is fixed at 1.2 m and the same distance (45 m) from sensor to point A and B, where the ball is dropped, is used. The difference between A and B is just the angle from the reference frame of NS-EW. The signal response is shown in Fig. 14, and the details are compared in Fig. 15 and Fig. 16. The apparent difference between signal A and signal B can be seen. Thus, it seems possible to obtain DOA by using the signal component of NS-EW and a proper algorithm.