Error Correction of Magnetic Gradient Tensor Localization for UXO-Like Targets With Portable Transient Electromagnetic Systems

The magnetic gradient tensor (MGT) algorithm, which can quickly locate underground targets without iteration, is constrained by positioning errors when detecting unexploded ordnance (UXO) with a transient electromagnetic (TEM) system due to the large sensor baseline distance. An innovative method based on a cross-shaped portable TEM system is proposed to improve the positioning accuracy of MGT localization. First, the MGT algorithm is introduced to estimate target positions and positioning errors. Second, an error correction model is constructed, with its parameters fit using the least squares algorithm. Finally, the estimated target position obtained from the MGT algorithm is corrected based on the error correction model. Experimental results show that the proposed method can notably improve the positioning accuracy of underground targets, achieving a horizontal position error of no more than 10 cm and a depth position error of no more than 5 cm. The inversion time takes approximately 25 ms, providing a novel method for real-time detection and rapid positioning of UXO near the surface.