Optimization Of AWPSO Algorithm For Conical Source Transient Electromagnetic With Ramp Step Current Excitation

Based on the conical source transmitter,this paper studies its transient response characteristics,turn-off effects and nonlinear optimization inversion methods under the excitation of oblique step currents,and provides inversion interpretation for transient electromagnetic shallow detection tasks.For reference,it can also enrich the theoretical system of conical field source transient electromagnetic data inversion methods.The transient response delay characteristics of different turns and different apparent resistivity models are studied in the ideal uniform half-space model;at the same time,the apparent resistance of the single-turn coil and the multi-turn coil is compared with the single-turn coil and the multi-turn coil under the equal magnetic moment(386.2 A·m~2).Based on the error between the rate and the true value,the critical conditions for the early and late delays are analyzed.The results show that the relative error curves of the early and late apparent resistivity corresponding to the changes of the conical field source turns and the resistivity of the geological model approximately coincide,indicating that the?/r0 value division and the field corresponding to the early and late delays The number of turns of the source and the resistivity of the geological model are independent of the two factors.The early and late t/r0 critical values?/r0 met by the three transmitters under equal magnetic moments are approximately the same;in addition,the conical field source is more effective than multi-turn coils and single-turn coils.Early entry into the early and late stages shows that the conical field source device can generate the same transient electromagnetic field instead of single-turn large loops and multi-turn coils,and is more conducive to the application of various theoretical conversion formulas under late conditions.In homogeneous medium,the method of Duhamel integration and Fitterman research is used to calculate the conical field source slope step response,and the transient response is converted into early and late apparent resistivity and full-area apparent resistivity;explored The influence of the shut-off effect on the transient response of the layered medium when the parameters such as the cover layer thickness and apparent resistivity of the two three-layer geoelectric models change,and the slope step response is corrected by Fitterman's late conversion coefficient;The changes before and after correction of the oblique step transient response in the five-layer(KHK)layered model of the cone-shaped field source and the multi-turn coil transmitter under the equal magnetic moment are discussed,and the layer is obtained by the"smoke ring"inversion.Resistivity and apparent depth of the medium-like medium.The research results show that the slope step effect is differently affected by the thickness of the cover layer and the apparent resistivity changes,leading to distortion and distortion of the transient electromagnetic signal.The corrected transient response is more consistent with the theoretical step response curve.The influence of the jump current is significantly weakened.For the conventional inversion method of the transient electromagnetic cone-type field source device,the initial model is artificial and inefficient.Through the analysis and research of the particle swarm optimization algorithm(PSO)and the neural network algorithm(BP),an improved Adaptive weighted particle swarm optimization(AWPSO)algorithm based on the neural network algorithm Sigmoid activation function.A standard test function is used to perform a trial calculation on the algorithm,and a number of theoretical layered geological models are established to verify the algorithm.Finally,field experiments are carried out in areas with known geological data,and the measured data is corrected by a slope step.The calculation results show that the new and improved algorithm has the advantages of higher global search and optimization ability,fast convergence speed,high calculation accuracy and no initial model during the inversion process.The experimental results show that the inversion results of the measured data before and after the correction are high in density.The electrical detection results are consistent,which verifies the effectiveness of the inversion algorithm.At the same time,the inversion results of the data after the oblique step correction have higher accuracy in identifying low-resistance bodies,showing that the oblique step correction method has a certain effect.