Intelligent Sensing in Inverter-fed Induction Motors: Wavelet-based Symbolic Dynamic Analysis

Abstract

Wavelet transform allows adaptive usage of windows to extract pertinent information from sensor signals, and symbolic dynamic analysis provides coarse graining of the underlying information for enhanced computational speed and robustness of sensor-data-driven decision-making. These two concepts are synergistically combined for real-time intelligent sensing of faults whose signatures are small compared to coefficients of dominant frequencies in the signal. Feasibility of the proposed intelligent sensing method is demonstrated on an experimental apparatus for early detection of rotor bar breakage in an inverter-fed induction motor.

Keywords

• Wavelet Analysis
• Symbolic Analysis
• Induction Motors
• Fault Detection