The complexity involved with the phase unwrapping procedure, while performing the existing spectral analysis of surface waves (SASW) on the basis of two sensors, makes it difficult to automate and requires frequent manual judgment. As a result, this approach generally becomes tedious and may yield erroneous results. The multichannel analysis of surface waves (MASW) technique can resolve the problem of phase wrapping. However, the MASW technique normally requires a large number of closely spaced sensors, typically 24–48 or even more. We have developed a new method that is fast, accurate, and generally resolves the unwrapping of phase with the use of just two sensors, provided the signal-to-noise ratio remains high. In this approach, the unwrapping of the phase can be performed without any manual intervention and an automation of the process becomes feasible. A few examples, involving synthetic test data and surface-wave tests, have been tested to determine the efficacy of our approach. Comparisons of the results have been made with the corresponding solutions using existing SASW and MASW techniques.
Implementation of Machine Learning Algorithms in Spectral Analysis of Surface Waves (SASW) Inversion
