Search

CN-122021066-A - Full waveform transient electromagnetic fast forward modeling method for pseudo-random encoding source

CN122021066ACN 122021066 ACN122021066 ACN 122021066ACN-122021066-A

Abstract

The invention discloses a full waveform transient electromagnetic rapid forward modeling method aiming at a pseudorandom encoding source, belongs to the technical field of geophysical exploration numerical simulation, and solves the problems of poor stability, low calculation efficiency and easiness in generating non-physical artifacts in the traditional pseudorandom transient electromagnetic full waveform forward modeling. The method comprises the steps of input data preprocessing, step kernel integral function preprocessing, emission current piecewise linearization processing, full waveform response piecewise analysis calculation, full-period boundary condition control and result output, wherein numerical burrs at a current jump position are eliminated from a physical mechanism, the calculation efficiency is improved by more than 180 times compared with a traditional Gaussian integral method, the numerical stability is high, the universality is good, and the method can be widely suitable for full-waveform transient electromagnetic forward modeling of various pseudorandom encoding sources.

Inventors

  • LU KAILIANG
  • YUE JIANHUA
  • FAN YANAN

Assignees

  • 中国矿业大学

Dates

Publication Date
20260512
Application Date
20260409

Claims (7)

  1. 1. A full waveform transient electromagnetic fast forward method for a pseudorandom encoding source comprising the steps of: S1, acquiring a transient electromagnetic step response core corresponding to a target ground model and a discrete sampling sequence of pseudo-random emission current; s2, carrying out numerical integration on the step response kernel, constructing a discrete integration kernel sequence, and constructing a quick retrieval interpolation model of the discrete integration kernel sequence through a conformal interpolation algorithm; s3, segmenting a discrete sampling sequence of the pseudo-random emission current according to sampling intervals, performing linearization approximation on the emission current in each sampling interval, and calculating current derivatives in each sampling interval; S4, aiming at each observation time, based on a convolution principle of a linear time-invariant system, converting continuous convolution integral into an analytic form according to a sampling interval of the emission current in a segmented mode, and calculating full waveform electromagnetic response of the observation time by carrying out differential weighted summation on a discrete integral kernel sequence; S5, in the calculation process of the step S4, causality boundary control and time-window extrapolation compensation are carried out on the input independent variables of the discrete integral kernel sequence; and S6, traversing all observation moments, completing calculation of full-time full-waveform electromagnetic response, and outputting and storing pseudo-random transient electromagnetic full-waveform response data.
  2. 2. The method for full waveform transient electromagnetic fast forward modeling for pseudorandom encoding source of claim 1 wherein in step S1, the transient electromagnetic step response kernel is transient electromagnetic response excited by step current under the same-observation device and same-ground model, and the discrete sampling sequence of the pseudorandom emission current is Wherein For the moment of sampling, Is that Emission current amplitude, j=1, 2, for time instant, , Is the total number of sampling points.
  3. 3. The full waveform transient electromagnetic fast forward method for a pseudorandom encoding source of claim 1 wherein said step S2 comprises: S21, performing numerical integration on the step response kernel by adopting a trapezoid integration method to construct a discrete integration kernel sequence, wherein an integration function corresponding to the discrete integration kernel sequence is that Wherein Is a transient electromagnetic step response core under step current excitation in unit time, Is the time lag amount; s22, carrying out interpolation reconstruction on the discrete integral kernel sequence by adopting PCHIP conformal interpolation algorithm, and establishing an integral function Is provided.
  4. 4. The full waveform transient electromagnetic fast forward method for a pseudorandom encoding source of claim 3 wherein in said step S3, at each sampling interval In, pseudo-random emission current Approximately linear, such that the current derivative within this interval is constant, i.e.: Wherein, the Representing the current derivative at time j, And The current magnitudes at times j +1 and j are indicated respectively, Representing the sampling time interval.
  5. 5. The full waveform transient electromagnetic fast forward method for a pseudorandom encoding source of claim 3 wherein in step S4 the full waveform electromagnetic response is specifically formulated as: Wherein B (t) is the full waveform electromagnetic response at the observation time t, Is the total number of sampling points, And The current magnitudes at times j +1 and j are indicated respectively, The time interval of the sampling is indicated, Represents 0 to The cumulative response of the time of day, Represents 0 to Accumulated response of time of day.
  6. 6. The full waveform transient electromagnetic fast forward method for a pseudorandom encoding source of claim 3 wherein in step S5 said full time period boundary condition control comprises: s51, causal boundary control of time lag At the time, let the integral function =0, Conforming to causality of transient electromagnetic response; S52, time window extrapolation compensation, namely time lag amount In which For the maximum time length of the pre-computed step response kernel, for the integral function A power law extrapolation compensation or truncation process is employed.
  7. 7. The full waveform transient electromagnetic fast forward method for a pseudorandom encoding source of claim 1 wherein said pseudorandom transmit current comprises m-sequences, bipolar inversion sequences, gold sequences.

Description

Full waveform transient electromagnetic fast forward modeling method for pseudo-random encoding source Technical Field The invention belongs to the technical field of geophysical exploration numerical simulation, and particularly relates to a full waveform transient electromagnetic rapid forward modeling method aiming at a pseudorandom code source. Background Transient Electromagnetic Method (TEM) is a time domain electromagnetic exploration method which is widely applied in the field of geophysical exploration, and is widely applied to a plurality of fields such as mineral resource exploration, hydrogeological investigation, engineering geological exploration, urban underground space exploration, geological disaster early warning and the like by virtue of the advantages of sensitivity to low-resistance geological bodies, large detection depth, convenience in construction and the like. Full waveform transient electromagnetic exploration is one of the important development directions in the field in recent years, and by transmitting excitation signals of complex waveforms such as pseudo-random codes, multi-step and the like, full-time electromagnetic response from a transmission power-on period to a later turn-off period is synchronously acquired, so that the signal-to-noise ratio, longitudinal resolution and effective detection depth of exploration data can be remarkably improved. The full waveform forward modeling is a core foundation for full waveform transient electromagnetic data processing, inversion interpretation and observation system optimization design, and the calculation accuracy, stability and efficiency directly determine the application effect of the full waveform exploration technology. For a linear time-invariant system, an arbitrary transmit current waveformElectromagnetic response under excitationThe expression is: The traditional convolution value implementation mode mainly adopts a Gaussian integration method or a direct value integration method. However, in the pseudo-random transient electromagnetic full waveform forward scene, the method has technical defects which are difficult to overcome: First, there are multiple rapid jumps in the current waveform of a pseudo-randomly encoded signal (e.g., m-sequence, bipolar flip sequence, gold sequence, etc.), the symbol widths are different, the current changes are severe near the rising and falling edges, resulting in current derivatives Which appears as an ideal impulse or sharp spike near the flip point. The traditional Gaussian product method based on fixed nodes is extremely sensitive to sampling point distribution, numerical burrs and non-physical fluctuation are extremely easy to generate at the corresponding response peak value of current jump, even oscillation divergence occurs, and physical rationality and numerical stability of a forward result cannot be guaranteed. Second, in order to suppress the numerical error, the conventional method needs to greatly increase the number of nodes of the gaussian product, resulting in exponentially increasing the calculated amount and extremely low calculation efficiency. Taking the conventional pseudo-random waveform forward modeling as an example, the single-channel calculation time consumption of the conventional 15000-node Gaussian integration method can reach more than 16s, the requirement of large-scale full-waveform forward modeling and inversion iteration on the calculation performance is difficult to meet, and the efficient processing scene of the pseudo-random transient electromagnetic full-waveform data cannot be adapted. Thirdly, the existing partial improved forward method can only process simple transmitting waveforms such as single step and bipolar, cannot adaptively process full-time response calculation of multiple random current turning of a pseudo-random encoding source in an electrified period, is poor in universality, and is difficult to adapt to full-waveform forward requirements of complex pseudo-random encoding excitation. In summary, the prior art has no technical scheme capable of simultaneously solving the problems of the forward numerical stability, the calculation efficiency and the universality of the pseudo-random transient electromagnetic full waveform, and becomes a key bottleneck for restricting the popularization and the application of the full waveform transient electromagnetic exploration technology. Disclosure of Invention Aiming at the problems existing in the prior art, the invention provides a full waveform transient electromagnetic fast forward modeling method aiming at a pseudo-random encoding source, which can effectively solve the technical problems that in the conventional pseudo-random transient electromagnetic full waveform forward modeling, the numerical stability of a traditional convolution integral method at a current jump is poor, the calculation efficiency is low, non-physical artifacts are easy to generate, and a complex pseudo-random encodi