Adaptive Regularization Inversion Theory Of Conical Field Source Transient Electromagnetic

Transient electromagnetic small size device with the advantages of portability and flexibility,widely used in engineering tunnel prediction,engineering shallow detection,coal mine roadway,and city.The traditional winding method of multi-turn small loop device had strong mutual inductance and long turn-off time,resulting in a certain range of shallow detection blind area,which was not conducive to shallow detection.To this end,this paper is based on the conical field source launching device,which has carried out the ground onedimensional,inversion theory and application research.The research contents of this paper enrich the theoretical method of conical field source transient electromagnetic inversion,and the research results provide theoretical reference for the processing of measured data.Based on conical field source device,the transient response characteristics of conical field source with uniform half-space model,2 layers(D and G),3 layers(H and K)and 4layers(KH)are studied.The response characteristics of conical field source are analyzed by induction potential curves,late apparent resistivity-time curves and depth-late apparent resistivity curves.The calculated late apparent resistivity-time curves and depth-late apparent resistivity curves can better reflect the electrical changes of the local electrical models,the stratification effect is good,and the reflection is more sensitive to the low resistivity layer.The apparent resistivity of the whole region of the conical field source transient electromagnetic is calculated by translation algorithm,which verifies the effectiveness of translation algorithm.Under the condition of equal magnetic moment,the characteristics of transient response time of conical field source,multi-turn coil and singleturn loop are studied.The results show that:The conical field source and multi-turn coil with equal magnetic moment enter the late stage earlier than the single-turn loop,and the late apparent resistivity curves of conical field source and multi-turn coil almost coincide.Therefore,the conical field source device can produce the same transient electromagnetic field in place of the multi-turn coil and the single turns.To address the difficulty of selecting the regularization factor in transient electromagnetic regularization inversion,this paper adopts the CMD adaptive regularization algorithm for automatic adjustment,and applies this method to the conjugate gradient inversion of conical field source transient electromagnetic.The forward and inversion calculations are carried out for 4 typical geoelectric models,which are consistent with the smoke ring.The inversion is compared,and the one-dimensional model inversion profile is compared.The model trial calculation also compares the inversion results obtained by adaptive regularization inversion and L-curve method.Finally,a field experiment was carried out in the work area,the measured data was processed,and the inversion processing effect was compared with the CSAMT detection results.The research results show that the "smoke ring" inversion has poor layering effect,the adaptive regularization inversion depth is closer to the layer interface,which improves the identification ability of the target layer interface,and the layering effect is good;the adaptive regularization inversion apparent resistance The rate curve is in better agreement with the theoretical model.When the initial models are different,after the inversion iteration is terminated,the relative fitting difference of the adaptive regularization optimization inversion is small,indicating that the inversion method has little dependence on the initial model and good stability.The relative fitting difference of the adaptive regularization inversion results is not much different from the inversion results obtained by the L-curve method,but the adaptive algorithm has higher computational efficiency and saves time than the L-curve method.The processing results of the adaptive regularization inversion method are consistent with the known data,and can effectively reflect the physical property differences of underground geological bodies.