The TFPF Based On Improved L-DVV Method In Random Noise Attenuation For Seismic Exploration

Seismic exploration is an important means of resource exploration, as well asone of the main methods of the oil exploration, gas resources, solid coal resourcesexploration. The quality of seismic record is the foundation of seismic explorationengineering. Therefore, improving the signal-to-noise ratio is one of the mostimportant work in seismic exploration. With rarely and complex resources, the targetsof seismic exploration are developing to deeper, thinner layer and more complicatedgeological conditions on the ground. The processing of seismic exploration data ismaking "three highs"(high signal-to-noise ratio, high resolution, high accuracy)requirements, in which "three highs" targets, high signal-to-noise ratio is the mostimportant requirement.The time-frequency peak filtering method is a signal processing algorithm whichdispense with the amplitude of processed signal is real. It is an effective tool fornon-stationary signal and multi-component signal in random-noise attenuation. Thetime-frequency peak filtering method is summarized into two steps: first of all,modulated the noised signal, into the instantaneous of an analytical signal with unitamplitude, then extract the peak in the time-frequency distribution to estimate thesignal effectively. In recent years, the time-frequency peak method in seismic dataprocessing has achieved significant effect, which can suppress random noise andretain the signal effectively,at the same time improve signal-to-noise ratio. TFPFtechnology is estimated without deviation under the condition: signal’s amplitude isthe linear function of time, namely the nonlinearity is close to zero. In the real seismicdata, processed signal is not always completely linear. The window length of thetime-frequency distribution is the key to the conventional TFPF technology.Traditional time-frequency peak filtering method takes a single window. Whenapplying TFPF to process seismic exploration data, the demands of attenuatingrandom noise and retaining signals are conflicting, the traditional way of selecting afixed window length can hardly make a balance of noise attenuating and signalretaining. This article puts forward a method, based on improved L-DVV to measurethe nonlinear for selecting window. Ideological line is: the signal is treated assubsection processing,then measure the degree of nonlinear. When most componentsof the part are signals, it is judged as signal part; the same with the judgment of thenoise part. Then use short window for the signal part, long for the noise part, andcombine the two parts as the final result. Compared with the single window lengthfiltering, the algorithm in this paper can meet the different demands of signal and noise, it has better result of noise reduction, meanwhile details of the signals are wellretained. Compared with point by point varying-window length TFPF, it has avoidedthe shortcomings of the existing varying-window length TFPF method such as thelarge computationally expensive and signal distortion.In order to verify the effectiveness of this algorithm in seismic data processing,first verify the accuracy of measuring nonlinear sequence by the improved L-DVVmethod. The method of improvement of L-DVV can not only distinguish andperformance the nonlinear more clearly for different parts of one sequence, but also inthe time series which contains strong impulse noise, the improved method of L-DVVcan still correct the degree of nonlinear for sequences. According to the rule that fromsimple to complex condition, the paper firstly applied improved TFPF to the syntheticseismic data, then to actual seismic data. Including single ricker wavelet，single phaseaxis,the synthesis of multiple phase axis and the actual earthquake records. Then,make comparison and analysis according to the waveform. The experiment resultshows that compared with the traditional single window length TFPF, the method candivide signal part and noise part reasonably, and create favorable conditions forchoosing time window length flexibly. Because of the combination of advantages oflong and short window length, both the preservation of the signals and reduction ofnoise have obtained good results.