| Surface Nuclear Magnetic Resonance(SNMR)technique is widely used in near surface exploration fields such as hydrological geological survey and disaster water source exploration due to its advantages of non-invasive direct detection and quantitative assessment of groundwater.However,due to the uneven spatial and temporal distribution of the Earth’s magnetic field and its susceptibility to magnetic interference and strong environmental noise,it is difficult to obtain accurate Larmor frequency in actual measurement.Therefore,the phenomenon of off-resonance is widespread,causing changes in signal amplitude and phase,leading to errors in hydrological information interpretation and reduced accuracy of deep detection.In response to the technical bottleneck of difficult to achieve effective resonance detection in the case of unknown offset of the Larmor frequency,scholars at home and abroad have conducted a large amount of research on biased resonance,which is still limited by the dependence on the true value of the Larmor frequency.Especially when the underground resistivity is low,the results of forward modeling and inversion interpretation are more severely affected by frequency bias.Therefore,how to improve the accuracy of off-resonance detection and the resolution ability of deep water bearing structures is a new direction for the development of this technique.In response to the above issues,this article proposes a dual-frequency off-resonance detection technique,which utilizes the same offset positive and negative transmitting frequencies on both sides of the Larmor frequency estimation value to excite the off-resonance,and uses this frequency offset to calculate the kernel function for inversion interpretation,solving the problem of detecting the off-resonance when the Larmor frequency estimation is inaccurate.However,lower underground resistivity can introduce significant phase shifts with depth in the dual-frequency off-resonance response signal,posing significant challenges to inversion interpretation and deep exploration.Therefore,this article establishes a comprehensive dual-frequency off-resonance detection model under the influence of multiple factors,develops a high-resolution imaging method that adapts to unknown Larmor frequency offset conditions,studies the response mechanism of dual-frequency off-resonance signals,proposes a dual-frequency off-resonance forward simulation and inversion algorithm based on resistivity phase correction,and achieves accurate imaging considering underground resistivity,Breaking through the bottleneck of the current SNMR detection of hydrological information interpretation accuracy constrained by the Larmor frequency.The effectiveness and reliability of the relevant research content have been verified through field measurement results.The details are as follows:(1)A dual-frequency off-resonance detection method is proposed to address the difficulty of effective on-resonance detection when the true value of Larmor frequency is unknown.Based on the theory of off-resonance detection,the approximate relationship of chamfering was derived,and the principle of dual-frequency off-resonance detection was theoretically proven.The core mechanism for solving the influence of unknown Larmor frequency offset was determined,laying a theoretical foundation for improving the accuracy of forward modeling and optimizing methods.In response to the establishment of a theoretical model and signal feature analysis for dual-frequency off-resonance,the response expression and forward modeling method for dual-frequency off-resonance signals were derived and studied.(2)In response to the low underground resistivity situation,additional phase acquisition causes low resolution and poor imaging accuracy in deep exploration using dual-frequency off-resonance.Combining the theory of dual-frequency offresonance and the generation mechanism of resistivity phase offset,a phase correction method combining resistivity and excitation magnetic moment in the kernel function for inversion is proposed,and the inversion kernel function expression of the dualfrequency off-resonance method in conductive media is derived,Effectively improving the inversion imaging accuracy of dual-frequency off-resonance data and expanding the application range of dual-frequency off-resonance method.(3)Aiming at the problem that the dual-frequency off-resonance data is mainly affected by power-line harmonic noise and random noise,the matching pursuit optimized by genetic algorithm data denoising method and the variational mode decomposition data denoising method are proposed.(4)Based on the traditional SNMR complex envelope inversion framework,a dual-frequency off-resonance inversion method based on L-BFGS algorithm is proposed.This method can fully combine the characteristics of dual-frequency off-resonance detection and complex envelope inversion,improving the detection depth and inversion accuracy.Aiming at the problems of slow solving speed and low iteration efficiency of conjugate gradient optimization algorithm in traditional inversion calculation,the latest L-BFGS algorithm is introduced,which greatly improves the inversion efficiency on the premise of ensuring the imaging accuracy,and realizes two-dimensional fast and accurate imaging of dual-frequency off-resonance.By simulating a one-dimensional layered model and a two-dimensional irregular water bearing structure,the effects of multiple factors such as unknown frequency offset,underground resistivity,set frequency offset,noise level,and transmitting time on the accuracy of imaging results were analyzed.The effectiveness of the dual-frequency off-resonance imaging method has been verified,achieving precise positioning and high-precision imaging of water bearing structures.(5)The dual-frequency off-resonance detection method proposed in this article was used to conduct field experiments in the suburbs of Changchun City,and compared with known drilling data to verify the effectiveness and accuracy of the dual-frequency off-resonance detection method.The dual-frequency off-resonance method can effectively detect situations where Larmor frequency estimation is not accurate,and the interpretation results are consistent with the drilling situation.In summary,the research content of this article further enhances the reliability and practicality of SNMR technique,which is of great significance for the further expansion and application of this technique. |
PDF Full Text Request