Regression Coefficients Suppress Early Signal Noise in Helicopter Transient Electromagnetic System

Abstract

Due to the increasing demand for mineral resources, mineral exploration is advancing toward difficult terrain areas. When a helicopter transient electromagnetic (HTEM) system is operating in complex terrain areas, severe-induced voltage noise occurs in the early signals, which affects shallow detection. Through analysis, we found that this noise essentially comes from the slight variation in the relative position between the receiving coil and the transmitting coil, resulting in a change in mutual inductance between the two. This change causes the primary transient electromagnetic (TEM) field to change constantly and affects observations of the secondary TEM field. Frequency amplitude analysis of actual early signals shows that the noise frequency can reach 0.06 Hz, which is lower than the 25-Hz excitation frequency. Therefore, based on the assumption that the mutual inductance remains constant within each half-cycle of excitation, this article proposes a method to suppress early signal noise by calculating the regression coefficient vector of high-altitude data. This method only uses induced voltage data to suppress noise and has a short processing time. Actual exploration line results verify that the peak value of the earliest gate signal decreases from 1856.9 to 155.12 nT/s, with an early observation time of $104.67 \; \mu $ s.