Sparse optimization for nonlinear group delay mode estimation.

Area of Science:

• Signal Processing
• Mechanical Engineering
• Applied Physics

Background:

• Nonlinear group delay signals with frequency-varying characteristics are prevalent in structural health monitoring and fault diagnosis.
• These signals often comprise multiple, potentially overlapping modes in the frequency domain, complicating analysis.
• Accurate decomposition and time-frequency representation of these modes remain a significant challenge.

Purpose of the Study:

• To develop a novel technique for estimating nonlinear group delay modes.
• To address the challenge of decomposing signals with overlapping frequency modes.
• To demonstrate the method's applicability in structural health monitoring and fault diagnosis.

Main Methods:

• Frequency-domain modeling of nonlinear group delay signals.
• Utilizing signal sparsity to develop a nonlinear group delay mode estimation technique.
• Employing an augmented alternating direction multiplier method for efficient implementation.

Main Results:

• The proposed method effectively estimates nonlinear group delay modes, handling crossed modes and transient impulse signals.
• Numerical simulations and experimental data validated the method's effectiveness and advantages.
• The technique shows potential for analyzing Lamb waves and bearing signals.

Conclusions:

• The developed nonlinear group delay mode estimation technique offers an effective solution for complex signal decomposition.
• The method has significant implications for advancing structural health monitoring and fault diagnosis applications.
• The efficient implementation facilitates practical application in real-world scenarios.