Three-dimensional Time-domain Electromagnetic Response Numerical Simulation And Intelligent Identification Of Polarization Media

China is rich in metal mineral resources,however,due to the low detection of mineral resources,the supply of mineral resources is far less than the demand,which will slow the development of industry.Achieving precise exploration of metal minerals is an effective way to further enhance the output of mineral resources under the new situation.Measuring the chargeability and conductivity effectively is an important method for detecting metal mines.The time-domain electromagnetic method is sensitive to low-resistance media,and has large detection depth and high accuracy.It has been widely used in the geological detection.Studying the magnetic-source time-domain electromagnetic induction-polarization combined detection method is essential to the high-precision measurement of metal mines.At present,the traditional concept believes that under step current excitation,the polarization response is only generated by the secondary induced current;under the ramp step current excitation,the polarization response is only generated by the primary induced electromotive force.However,the production process of the polarization response in the actual measurement is more complicated,and the production mechanism of the polarization response remains unclear.Carrying out the numerical simulation of the three-dimensional electromagnetic response of the induction-polarization symbiosis effect is an important means to understand the polarization characteristics.The introduction of the fractional Cole-Cole complex conductivity model makes it difficult to discretize the Ohm's law convolution,resulting in poor accuracy and low efficiency.In addition,the polarization response is easily affected by the earth structure and measurement parameters,and the negative response is not obvious.Thus,the responses are easy to be mistakenly identified as non-polarization responses,which leads to low interpretation accuracy and ultimately reduces the detection resolution of metal mines.In order to solve the above problems,the survey is supported by the key project of the National Natural Science Foundation of China "Research on key technology of electromagnetic detection for induction-polarization symbiosis effects of bio-phase conductive medium based on SQUID" and “Research on anomalous diffusion mechanism of time-domain electromagnetic detection based on fractional finite difference method”.Three-dimensional time-domain electromagnetic response numerical simulation and intelligent identification of polarization media are carried out.The main contents are as follows:(1)The analysis of the mechanism of magnetic-source induction-polarization symbiosis effect.The thesis unified the analysis method of time-domain electromagnetic field with the circuit analysis method,and then established the induction-polarization symbiosis circuit model.Under the ramp step waveform,the generation process of the polarization current was analyzed.To quantify the contribution of the primary field and the secondary field,the energy contribution rate was defined.The influence of the switch-off time on the excitation process was discussed.The generation process of the polarization current in classical media such as sulfide ore,graphite ore,frozen layer,etc.,has been studied to clarify the relationship between switch-off time,polarization parameters and excitation source.(2)The frequency domain rational function was used to approximate the fractional Cole-Cole conductivity model.The absolute value relationship was used to transform the nonlinear problem into a linear programming problem to simplify the solution process.Based on the properties of the same base power function,the integral variable and the time variable were separated,and the trapezoidal integration method was used to construct the discrete recursive formula of Ohm's law,which solved the problem of the discreteness of Ohm's law and improved the calculation accuracy and efficiency.(3)Intelligent recognition of IP responses based on PMI-FSVM algorithm.The thesis defined the polarization impact ratio to quantify the strong and weak IP effects by considering the effects of topological structure of the earth,polarization parameters,and measurement system.The negative slope fitting based on piece-wise nonlinear least square method,an optimal feature selection method based on the partial mutual information,and an IP recognition method based on the fuzzy support vector machine were proposed.The partial mutual information algorithm screened two parameters,the sign reversal time and the slope value of the last time,which were the most critical to characterize the IP effect.The non-linear indivisible data was mapped to the high-order spatial dimension through the radial basis function to realize the identification of IP effects.The accuracy of the recognition of polarization response was over 90%.The one-dimensional lateral constrained inversion theory was used to extract and interpret the polarization responses.(4)Field measurement and identification algorithm verification.Based on the actual measurement situation,the causes of the "fake" polarization phenomenon were analyzed in detail,and then were eliminated.Based on the induction-polarization symbiosis effect model,the polarizable anomaly loop was designed.The electromagnetic response curves of the polarization medium were simulated using polarizable anomaly loop in field measurement,and the accuracy of the identification algorithm was verified.Based on the measurement data of field experiments in the Siziwang Banner of Inner Mongolia and Denmark,the intelligent identification of polarization effects and the interpretation of parameters such as resistivity and chargeability were carried out,and the effectiveness of the algorithm was further verified by comparing with geological data.This paper realizes the three-dimensional time-domain electromagnetic response numerical simulation and intelligent identification of polarization media,and proposes innovative ideas in the study of basic principles,numerical simulation,actual detection,data recognition and interpretation,etc.,which lays a foundation for the realization of high-precision detection of metal mines.Basic,it has good portability and universality,and can also be used in other electrochemical effects and other fields.