A novel threshold segmentation instantaneous frequency calculation approach for fault diagnosis

Area of Science:

• Signal Processing
• Mechanical Engineering
• Vibration Analysis

Background:

• Instantaneous frequency is crucial for analyzing non-stationary signals.
• Existing methods can be influenced by amplitude mutations.
• Accurate analysis is vital for applications like fault diagnosis.

Purpose of the Study:

• To propose a new method for calculating instantaneous frequency.
• To address the challenges posed by amplitude mutations in non-stationary signals.
• To enhance the accuracy of fault diagnosis in rolling bearings.

Main Methods:

• A threshold segmentation instantaneous frequency calculation method is proposed.
• The method compares single component signals with Second Order Differential Equations in polar coordinates.
• Segmentation is achieved by detecting amplitude mutation points based on signal area changes.

Main Results:

• The proposed method effectively segments non-stationary signals.
• It reduces the influence of amplitude mutations on instantaneous frequency calculations.
• Simulations, derivations, and experiments validate the method's performance.

Conclusions:

• The developed method is computationally simple and robust.
• It accurately captures transient signal information.
• The method demonstrates effectiveness in rolling bearing fault diagnosis.