Research On Seismic Data Acquisition And Noise Suppression Method

As a way of Geophysical Prospecting, seismic exploration plays an important role in oilfield development. Field seismic data acquisition and signal processing are the two main aspects in seismic exploration. Only high-quality seismic section can provide a reliable basis for oil exploration. The real field data collection process is often suffered all kinds of noises such as regular interference waves and random interference waves. As a result, the gathered data can’t be used for seismic inversion and explanation directly. To provide a theoretical basis in equipment selection, this paper analyzes the modulator’s quantization noise in seismic exploration instrument. Considering the high-speed data output stream in modulator, we design a decimation filter based on FPGA. A new noise reduction algorithm is put forward aiming at reducing the random noise in seismic exploration. The main contents are explained as follows:1) Set up the linear equivalent model of Sigma-delta ADC modulator. The quantization SNR of first-order and second-order modulator is analyzed in detail. The relationship between quantization SNR and sampling rate of high-order is deduced. The results show that the fourth-order Sigma-delta ADC modulator’s quantization SNR is more than 120 d B. It can meet the demand in high-precision seismic exploration.2) Design a digital decimation filter based on FPGA with the hardware of xc5vlx85-1ff676. First, we choose the appropriate filter structure. Second, use the hardware description language of verilog and VHDL to edit the filter code. The FIR filter is designed with the help of IP core in ISE, which can improve the development process greatly and make full use of the chip resources. This digital decimation filter has two main functions, one is 512 times down-sampling and the other one is filtering out band noise.3) The characteristics of common noise in seismic exploration are introduced. Aiming at reducing the random noise in seismic exploration, we study the theory of wavelet threshold de-noising. Then, we propose a new threshold de-noising method. The experiment in MATLAB shows that this method can achieve a proper result in removing the random noise.