Primary Multiplicative Noise Suppression in the HTEM System Via Adaptive Momentum Estimation and Elastic-Net Regularized Framework

The bird’s movement causes a variable primary response in the helicopter transient electromagnetic (HTEM) system called primary multiplicative noise (PMN). PMN dominates measured data on time and early off time and will significantly impede the interpretation of shallow earth. Measuring and estimating PMN is difficult due to the complexity of the kinematic state and the high testing cost. We propose to use the elastic-net regularized linear regression (ENLR) framework to suppress PMN and adaptive momentum estimation (Adam) to speed up the process. First, we analyze the generation mechanism and time-frequency domain characteristics of PMN. Second, we built the ENLR framework and designed the loss function based on the attributes of PMN and signals. Then, we employ Adam to solve the loss function and obtain the estimated PMN of all half periods. Finally, the PMN is subtracted from the original signal. Field tests show that the proposed method effectively removes the PMN, advances the earliest time gate that meets the error requirement from 255.21 μs to 20.83 μs after the first time gate, and improves the system’s ability to detect anomalies as shallow as 20 m.