Transient electromagnetic one-dimensional inversion based on differential evolution algorithm

The transient electromagnetic（TEM）data collected in practice encompasses both electromagnetic induction and induced polarization（IP）effects. Accurately extracting information on resistivity and polarization is crucial in the interpretation of electrical source TEM data. Therefore，firstly，the forward modeling is achieved by the finite-length electric source TEM method with a Cole-Cole complex resistivity model. On this basis，a one-dimensional inversion method of electrical source TEM based on a differential evolution algorithm is proposed. Based on the traditional differential evolution algorithm，the reverse learning strategy and the adaptive adjustment of control parameters are introduced to accelerate the convergence of the inversion. Meanwhile，constraint conditions are introduced into the objective function to form the minimum structure inversion，which reduces the multi-solution of the inversion. Based on the typical three-layer geoelectric model and complex multilayer model，the theoretical model is tested，and the resistivity and polarization of the model can be effectively restored by the inversion results. Finally，the measured data are used for inversion，and the inversion resistivity is consistent with that obtained by OCCAM. On the basis of the resistivity constraint，the polarization information is obtained by further inversion. Based on this resistivity constraint，further inversion is performed to obtain polarization information. The inversion results indicate that the algorithm proposed in this paper can accurately extract resistivity information from the measured data and obtain polarization distribution of underground media. It demonstrates the accuracy and applicability of the algorithm. © 2024 Editorial office of Oil Geophysical Prospecting. All rights reserved.