Intelligent Ball Bearing Fault Diagnosis Using Fractional Lorenz Chaos Extension Detection

Area of Science:

• Nonlinear Dynamics and Chaos Theory
• Mechanical Engineering and Condition Monitoring
• Signal Processing and Fault Diagnosis

Background:

• Ball bearing failures can lead to significant machinery downtime and safety risks.
• Traditional condition monitoring methods may lack sensitivity or be computationally intensive.
• Chaos theory offers potential for detecting subtle anomalies in vibration signals.

Purpose of the Study:

• To develop an efficient and inexpensive method for ball bearing condition monitoring.
• To investigate the application of fractional-order chaotic systems and extension theory for fault detection.
• To compare the proposed method with existing signal analysis techniques.

Main Methods:

• Utilized a non-autonomous Chua's circuit and a fractional Lorenz chaos system as master and slave systems.
• Introduced bearing vibration signals, measured by acceleration sensors, into the chaotic systems.
• Applied extension theory, specifically the matter-element model, to analyze dynamic errors and determine bearing condition.

Main Results:

• The fractional-order master-slave chaotic system achieved a 100% diagnostic rate for bearing faults with parameter adjustment.
• Extension theory effectively established classical and sectional domains for fault condition identification.
• The proposed method demonstrated superior or comparable performance to discrete Fourier transform, wavelet analysis, and integer-order chaos systems.

Conclusions:

• Fractional-order chaotic systems combined with extension theory provide a highly accurate and efficient approach for ball bearing monitoring.
• This method offers a cost-effective solution for real-time condition monitoring applications.
• The study highlights the potential of chaos-based signal processing for mechanical fault diagnosis.