A Novel Fault Detection and Identification Framework for Rotating Machinery Using Residual Current Spectrum

Area of Science:

• Engineering
• Mechanical Engineering
• Electrical Engineering

Background:

• Rotating machinery, particularly induction motors (IM), are critical in industrial applications.
• Effective fault detection and identification (FDI) are essential for maintaining operational integrity and preventing failures.
• Existing FDI methods may face challenges in identifying diverse or unknown fault signatures.

Purpose of the Study:

• To propose a novel model-based fault detection and identification (MFDI) framework for induction motor (IM)-driven rotating machinery (RM).
• To enhance frequency-domain FDI by utilizing a residual signal derived from a data-driven subspace identification (SID) model.
• To validate the proposed FDI framework's efficacy in an industrial setting, including diagnosing unknown fault signatures.

Main Methods:

• Employed a data-driven subspace identification (SID) algorithm to derive the IM state-space model from voltage and current signals under quasi-steady-state conditions.
• Developed a novel frequency-domain FDI approach by replacing the current signal with a residual signal, applying thresholding for analysis.
• Utilized the statistical Q-function to define fault frequency bands, distinguishing fault signatures from noise.

Main Results:

• The residual spectrum proved more sensitive to fault signatures compared to the traditional current spectrum.
• The proposed FDI framework successfully identified fault signatures in experimental tests on a wastewater pump.
• The system demonstrated capability in diagnosing both mathematically known and unknown faulty signatures.

Conclusions:

• The novel MFDI framework offers improved sensitivity and accuracy in detecting induction motor faults.
• The residual signal-based approach enhances the identification of fault signatures, particularly for unknown fault types.
• The successful industrial validation confirms the practical applicability of the proposed FDI method for rotating machinery.