A Method for Detecting Dynamic Mutation of Complex Systems Using Improved Information Entropy.

Area of Science:

• Nonlinear Dynamics
• Signal Processing
• Mechanical Engineering

Background:

• Time series analysis is crucial for understanding complex systems.
• Existing entropy methods may lack sensitivity or require extensive parameter tuning.
• Rolling bearing fault diagnosis demands robust and efficient analysis techniques.

Purpose of the Study:

• To introduce a novel nonlinear analysis method, improved information entropy (IIE).
• To evaluate the efficacy of IIE in analyzing time series complexity and distinguishing signals.
• To apply IIE for sensitive fault diagnosis in rolling bearings.

Main Methods:

• Developed a special probability mass function for normalized Shannon entropy calculation.
• Applied the improved information entropy (IIE) method to various typical time series.
• Validated IIE's performance through experimental fault diagnosis of rolling bearings.

Main Results:

• IIE effectively distinguishes different signal types and reflects time series complexity.
• The method demonstrates high sensitivity to signal changes.
• Experimental results confirm IIE's ability to extract sensitive bearing running state characteristics.
• IIE offers fast operation times and minimal parameter requirements.

Conclusions:

• Improved information entropy (IIE) is a sensitive and effective nonlinear analysis tool.
• IIE provides a robust approach for rolling bearing fault diagnosis.
• The method's efficiency and low parameter dependency make it practical for real-world applications.