Three-Dimensional Inversion of Semi-Airborne Transient Electromagnetic Data Based on a Particle Swarm Optimization-Gradient Descent Algorithm

Abstract

Semi-airborne transient electromagnetics (SATEM) is a geophysical survey tool known for its ability to perform three-dimensional (3D) observations and collect high-density data in large volumes. However, SATEM data processing is presently restricted to 3D model-driven inversion, which is not conducive to detailed surveys. This paper presents a new 3D model- and data-driven inversion algorithm using the particle swarm optimization (PSO) and gradient descent (GD) algorithms. PSO is used to suppress the multiplicity of solutions associated with inverse problems, and the GD algorithm is employed to accelerate the convergence of the inversion process. For the PSO-GD algorithm, a new model-updating equation is established and a cosine probability function is introduced as a weighting term for PSO and GD algorithms to ensure a smooth transition between the two algorithms in the iterative process. The α-trimmed filter function is used as a regularization constraint to smooth the model. The stability and reliability of the PSO-GD algorithm are verified through numerical simulations. Finally, the new algorithm is applied to the processing of SATEM measurements of the Qinshui coal mine in Jincheng, Shanxi Province, China.