Joint Inversion Of One-Dimensional Time-Frequency Electromagnetic Data With Constraints

The time-frequency electromagnetic method is a new type of artificial field source electromagnetic method proposed by optimization and innovation based on the controllable source audio magnetotelluric sounding method and the long-offset transient electromagnetic method.This method can simultaneously obtain frequency domain and time domain electromagnetic information through one signal acquisition.Compared with a single frequency domain or time domain electromagnetic signal,the time-frequency electromagnetic method comprehensively utilizes frequency domain and time domain electromagnetic information,which can further improve the exploration accuracy.This method has been widely used in oil and gas,shale gas exploration and other fields.In order to fully exploit the time-frequency electromagnetic information and obtain a refined geoelectric model,this paper carries out a constrained time-frequency electromagnetic joint inversion study.Based on the regularization inversion theory,this paper constructs a onedimensional adaptive time-frequency electromagnetic joint inversion algorithm.The algorithm introduces a variety of model constraint schemes into the objective function.At the same time,considering the data magnitude,an adaptive balance factor is added to the frequency domain and time domain data.The adaptive regularization scheme is adapted and optimized for the inversion algorithm.At the same time,it is proposed to integrate the model parameter search range constraint with the step size constraint scheme to avoid the influence of the empirical value of the traditional method.When the parameter search range is limited,the inversion iteration step size is synchronously constrained to ensure the stability of the inversion process.Several theoretical models are designed to test the algorithm,and the feasibility and superiority of the proposed inversion algorithm are verified by controlling the regularization factor,data trade-off factor and parameter constraint scheme.The inversion algorithm can adaptively adjust the regularization factor and data trade-off factor in the inversion iteration process,and adjust the inversion iteration step size through the feasible region and WOLFE criterion based on the reliable resistivity prior information,so as to quickly and accurately identify the target layer resistivity and layer thickness information.On this basis,the typical geoelectric model,theoretical model with different noise levels and measured data are used for inversion test.The test shows that the new scheme of time-frequency electromagnetic joint inversion proposed in this paper can effectively improve the resolution of resistivity inversion results of complex models,and its inversion effect is obviously better than that of single data type.The inversion scheme has strong ability to identify oil and gas and shale gas reservoirs,and has certain application potential in the field of time-frequency electromagnetic exploration.This study can enrich the theoretical method of time-frequency electromagnetic data inversion,provide theoretical reference for the processing and interpretation of measured data,and has practical significance for improving the recognition accuracy of time-frequency electromagnetic methods for oil and gas reservoirs.