Propagational Characteristics For Seismic Wave And Study On The Chaotic Vibrator Detection Method For Weak Signal In Seismic Prospecting

In the recent years, with the increasing demand for oil and gas, theprospecting horizon has turned to develop deep earth, thus the forecasting of oiland gas reservoirs is closer to lithosphere and deep earth, and tectonics andevolution of deep earth directly influence the generation, storage and movement.Simultaneously, people pay more attention to the seismic prospectingtechnologies. Extracting weak effective signals from strong noise is a crucialcontent of modern information theory. Traditionally, a signal is undetected if thesignal is weaker than noise. Nevertheless, weak signal detection techniquefounds a novel approach to extract signals from strong noise and makes thedetectable threshold value lower than noise. Detecting effective signals (such asharmonic wave, periodic pulse, etc) under the background of strong noise, thechaotic vibrator detection (CSD) method is a new stage to the development ofweak signal detection technique.In the seismic prospecting, for the record with SNR<0dB, it is still a puzzlehow to identify and detect seismic events. Generally speaking, the larger thedifference in the impedance between up and down reflecting layer is, thestronger the corresponding reflected information. The less difference ofimpedance and distance attenuation will make received reflected informationmuch weaker. Under the background of strong noise, although the reflectedinformation is strong, it will be unconspicuous even invisible in the record. As toexisting but unidentified effective reflected events, it is required to introducecertain technique to detect them. Therefore, in theory and practice, combiningwith the kinematic and dynamic natures of seismic wave, the paper detaileddissertates the approach, procedure and problems how to utilize the CSD methodto detect weak signals in seismic prospecting. The fundamental theory of theCSD is to utilize the mutational characteristic in the phase-state of system,namely, the transition from chaotic phase-state to periodic phase-state, to identifyeffective signals.Based on the summarization of status quo for SNR enhancement in seismicprospecting, the analysis of relevant methods enhancing SNR, and the briefdescription of developing history for chaotic theory and signal detectingapplication, the paper mainly carries out further studies from below four aspects.As a result, we obtain much new cognition establishing a certain fundament onthe application in the weak signal detection using the CSD method in seismicprospecting.1. Microscopic characteristics of elastic wave propagation in solid mediaFrom the view of quantum mechanics, the chapter studies the microscopiccharacteristics of elastic wave propagation in solid media, which establish afoundation to deeply understand the propagation of elastic wave in solid wave.And then the cellular-automata finite deep potential trough model is furtherextended. The paper obtains the relation between the vibratory velocity ofmolecular group (MG) and physical wave velocity, and then deduces the relationsatisfied by the MG's energy through solving the Schrodinger equation instationary state.When studying the MG's energy, the paper presents two variablescharacterizing microscopic and macroscopic qualities of MG, namely, the cut-offradius coefficient μ that characterizes microscopic scale and the equilibriumradius coefficient τ that characterizes macroscopic scale, which indicates thatthe MG has the qualities both microscope and macroscope. Furthermore, byLennard-Jones(12-6) potential function and introducing the coefficients ofrepulsion force V12 and attractive force V6, the paper presents the potentialfunction fit for solid media, and obtains the variable relation of V12/ V6 followedby μ and τ , namely, with the augmentation of μ V12/ V6 increases, but withthe augmentation of τ V12/ V6 decreases, accordingly, the paper obtains thepotential function of MG, i.e. Lennard-Jones(10-6) potential, and then illustratesthe potential distribution in the model that shows periodic "barrel" shape.After discussing the potential of model, the paper acquires some possibledistribution and regularity on the MG's energy. The MG's energy is providedwith the structure of continuous energy belt (when U(R)=0) and discontinuousenergy belt (when U(R)≠0) and the interval between adjacent energy is graduallyenlarged. When U(R)=0, the MG's energy shows the particle's energy mode inthe infinite deep potential trough, but the quantum number n can be 0 in themodel of the paper, what's more, when n=0 the energy magnitude is directproportion to μ , whereas inverse proportion when n equals other values. WhenU(R)≠0, by the equivalent relation between MG's energy and potential, the paperexpatiates the MG's energy distributions corresponding to different τ , whichhave extreme difference. In addition, the paper deduces that the MG's wavelength has the order of 10-20m.2. Fundament detecting weak signals using chaotic vibrator systemBased on the analysis of chaotic information technology and development,by studying the origin of nonlinear resilience force item and nonlinearconservative system, combining the phase-states of chaotic system with differentnonlinear resilience force items when given initial value, and by math deductionand validation, the chapter presents the selecting rules to the nonlinear resilienceforce item of system, namely, the coefficient and power of item with maximumexponent in the Potential-Hamiltonian equation must satisfy a N>0 and N isodd number, respectively, what's more, there are not strict limitation for thenumber of items, and positive or negative of coefficients and even or odd ofpowers for the former N-1 items within f(x). This rule complements the usablescope of chaotic system used to detect signals.In succession, based on the analysis of intuitionistic and quantified methodsfor chaotic criterions, the paper presents the united approaches between thePoincare section with "Periodicity 3" and the Lyapunov exponent (LE) spectrumand LE time evolutional curve to confirm the threshold value from the chaoticphase-state to periodic phase-state of the system. By contrasting and analyzingthe LE spectrums with different nonlinear resilience force items, the paper pointsout that we must consider the "stability" and "mutation" before and after thecritical phase-state of system besides the selecting rules for the nonlinearresilience force items, when selecting the better system suited to detect weaksignal. The stability is the length of time keeping the chaotic phase-state ( λ1 >0)before the critical phase-state and the periodic phase-state ( λ1 <0)after thecritical phase-state, and the longer the length, the better the stability. Themutation is the difference of λ1 values between before the critical point andafter the critical point corresponding to the system, and the larger the difference,the stronger the mutation. The good stability can ensure the stability of detectedcritical threshold value and the strong mutation can ensure the sensitivity ofsystem to weak signals.3. Influence of noise to the CSDBy describing the method and procedure detecting a harmonic signal usingthe CSD, the chapter detailed discusses the influence of noise to the CSD in theprocedure using the Poincare section with "Periodicity 3", LE spectrum and LEtime evolutional curve.In general, the augment of noise power will result in phase transition tochaotic phase-state and make the stability worse and the mutation weakerwhenever the system locates in periodic or nonperiodic phase-states. As shown inthe LE spectrum, with the augment of noise power, the minimum value of λ1before the critical phase-state is increasing, which is the same tendency to λ1after the critical phase-state. Simultaneously, as shown in the Poincare, with theaugment of noise power, the mapped points change from centralized point set inthree fixed regions with a certain direction to unorderly dispersed point set in thewhole mapped region.Moreover, the augment of noise power will destroy the periodic bifurcationof system. When noise is not added, the periodic bifurcation can be producedwith the variation of γ before the critical phase-state, but with the enlargementof noise power, the frequency of the periodic bifurcation is decreasing.In addition, when noise power is less, the system occurs λ1 <0, namely, thesystem reaches to a periodic phase-state, which shows the "Acceleration'function to periodic phase-state evolution. It is required to identify the functionbecause the CSD requires that the nonperiodic phase-state of system isunchanged once noise is added, i.e. λ1 is still more than zero, so we can judgethat the phase mutational transition is resulted from weak signal other than noisewhen noise and signal are added together.Similarly, we should ensure λ1 <0 when adding only noise, which cammake the system exhibit the stable passing belt, and then the system exhibit theunstable alternate belts before and after the stable passing belt. With theenlargement of noise power, the stable passing belt is shrunk as a whole, forexample, its interval scope, the detectable threshold value of the system, thethreshold value of SNR, and λ1 within the stable passing belt are all decreasing,which results in the decreasing weakening of the ability to detect weak signals bythe CSD.4. Detecting weak seismic event using the CSDThe chapter significantly dissertates the processing flow and key proceduresusing the CSD to detect weak signal (SNR<0dB) in seismic prospecting.Firstly, the paper illustrates the detailed procedure how to convert an eventin a common shot record to a quasi-periodic signal sequence, namely, the theoryand effect of the horizontal dynamic correction (H-DC). Using a group ofscanning velocities and times to scan and truncate the event in a syntheticcommon shot record, the paper analyzes the relocation and degree of completionof obtained wavelets after the H-DC, and by numerous calculation and figures,the paper validates that what location each obtained effective wavelet comesfrom the event can be confirmed by solving the relocating equation, combiningwith the correlation coefficient between the scanned wavelet and completewavelet, and how the degree of completion (waveform) of each scanned waveletcan be confirmed by calculating the sum of the absolute value of each scannedwavelet and then contrasting it with the waveform sequence figures, combiningwith the correlation coefficient between the scanned wavelet and completewavelet. Furthermore, the paper discusses the attentive problems in the scanningprocedure.Secondly, the paper founds the flow-chart detecting three kinds of commonshot records with SNR(∞,0dB,-3dB) by the CSD. A common shot seismicrecord can be inputted into the chaotic vibrator system after several processingstages, namely, the selection of the chaotic system, frequency spectrum analysis,prefiltering, testification of parameter groups, scanning and moveout in variablevelocity and variable time-window according to each arrival time, detection inthe chaotic system, auto interpretation to results, general processing, and errorestimation and explanation, etc. Moreover, the chapter analyzes the relevantproblems, such as SNR, scanning time interval and noise when occurring in theflow-chart.Thirdly, based on the analysis of distorting category for seismic wavelet, thepaper detailed depicts the procedure and effect detecting the distorted Rickerwavelet sequence using the CSD. The paper realizes the detection to the distortedRicker wavelet by judging λ1 shown in former chapter, and then obtains theratio of energy between the distorted quasi-periodic Ricker wavelet sequence andwhite noise is approximate 1:5.Fourthly, in order to eliminate the stochastic noise in the record, the papertakes the filtering method along trace one by one to a synthetic record with(SNR)1=0dB and an event. After the filtering processing, the SNR of the recordis enhanced to 16.77dB.In the end, the paper summarizes the main results, and then analyzes anddiscusses the further issues required to study on the CSD method for weak signalin seismic prospecting and other applications of the CSD. Although theapplication of CSD in seismic prospecting is still in the preliminary stage, theCSD innovates a novel approach to detect weak effective information in seismicprospecting, and it can be an effective method to contribute to the forecasting ofoil and gas and tectonics of inner earth.