A Signal Reconstruction Method for Memory-Type Transient Electromagnetic Detection Systems in Horizontal Wells

Abstract

A signal reconstruction method for memory-type transient electromagnetic detection systems based on empirical mode decomposition (EMD) was proposed in this study to tackle with the large volume of data saved by the electromagnetic flaw detector in horizontal wells. First, based on the underground transient electromagnetic detection model, sparse representation was performed for the depth-dimensional transient electromagnetic detection signal, which was then projected through a Gaussian random measurement matrix. Next, a depth-directional observation matrix was built to reconstruct the underground transient electromagnetic signal. In addition, in order to improve the reconstruction precision, the baseline wander of underground transient electromagnetic detection signal was removed through EMD, the key information of casing pipe was described using intrinsic mode functions (IMFs), and the IMFs were constructed using the above compressive sensing algorithm. It is verified through the measured data that the compressed signal after the baseline removal can be recovered at a higher probability, and moreover, it can accurately describe the forms of casing pipes and effectively remit the storage problem of large data volume faced by the memory-type electromagnetic flaw detectors in horizontal wells.