Abstract
Smart manufacturing systems require real-time, non-intrusive monitoring to ensure precision and reliability in machining operations. Traditional sensor-based methods, such as force or vibration monitoring, are often impractical in industrial settings due to their high cost, complexity, and sensitivity to environmental factors. This research addresses these challenges by advancing Motor Current Signature Analysis (MCSA) as a unified framework for both machine condition monitoring and process monitoring.The thesis introduces a novel analytical model of motor current and develops a comprehensive electromechanical coupling model that characterizes the dynamic behavior of motor currents in spindle systems, considering nonlinear mechanical vibrations and electromagnetic interactions. These models offer critical insight into how machining parameters—such as cutting parameters, tool wear, and belt tension—affect motor current signals through amplitude and phase modulation (AM and PM).
To improve the accuracy of feature extraction, this research proposes the Modulation Signal Bispectrum (MSB)method for demodulating AM and PM, along with the Squared Sideband MSB (SS-MSB) for quantifying the joint AM-PM (JAPM) effects under low signal-to-noise conditions. Additionally, a Carrier Reconstruction MSB (CR-MSB) technique is introduced to mitigate spectral distortion caused by closed-loop control (CLC) systems common in CNC machines.
The proposed models and methods are validated through simulation and experimental testing on turning and CNC milling platforms. Results demonstrate that these techniques can effectively detect variations in machining conditions and mechanical faults, such as gearbox faults, belt looseness, and transmission anomalies, by analyzing changes in current modulation patterns. This work provides a robust, data-driven solution for comprehensive machine and process monitoring, supporting predictive maintenance and precision manufacturing.
| Date of Award | 29 Apr 2025 |
|---|---|
| Original language | English |
| Supervisor | Fengshou Gu (Main Supervisor), Helen Miao (Co-Supervisor) & Wenhan Zeng (Co-Supervisor) |