Sparse and Dispersion-Based Matching Pursuit for Minimizing the Dispersion Effect Occurring when Using Guided Wave...

Area of Science:

• Materials Science
• Mechanical Engineering
• Signal Processing

Background:

• Ultrasonic guided waves are crucial for structural health monitoring (SHM) and defect detection.
• Guided wave signals often exhibit dispersion, multiple modes, and noise, complicating analysis.
• Existing signal processing methods struggle with these complexities, hindering accurate defect identification.

Purpose of the Study:

• To develop an advanced and robust signal processing tool for analyzing complex ultrasonic guided wave signals.
• To address challenges posed by signal dispersion, mode overlap, and noise in structural health monitoring.
• To enhance the accuracy and reliability of non-destructive testing (NDT) for defect detection.

Main Methods:

• Proposed Sparse Representation with Dispersion Based Matching Pursuit (SDMP) algorithm.
• Designed an overcomplete dictionary of basic atoms using Finite Element Method (FEM) simulations.
• Implemented a two-stage SDMP process involving atom selection based on time localization and frequency consistency.

Main Results:

• SDMP successfully separates overlapped wave modes and effectively suppresses noise.
• The algorithm demonstrates high sparsity and accurate signal approximation.
• Numerical simulations and experimental results on steel pipes validate SDMP's effectiveness for damage detection.

Conclusions:

• SDMP provides an effective and robust solution for analyzing complex guided wave signals in SHM.
• The method enhances defect detection capabilities by improving signal interpretation.
• SDMP offers a promising advancement in non-destructive testing technologies.