Guided Wave NDT Signal Recognition with Orthogonal Matching Pursuit Based on Modified Evolutionary Programming

This work aims to decompose and construct guided wave NDT signal by orthogonal matching pursuit employ modified evolutionary programming algorithm based on the t distribution (MtEP), which can reduce the computation complexity. A pipe sample with holes is inspected by guided wave and the test signal is processed by the OMP with MtEP. The processed result is compared with the OMP with evolutionary programming algorithm based on the t distribution (tEP). The result shows that this improved method can reconstruct the signal with better quality, less calculation time, and the echoes signal can be recognized clearly. Therefore, this proposed approach is useful to the guided wave NDT signal recognition.