A new fault indicator for reducing calculation time and digital system memory size: Induction motor application

This paper studies broken rotor bar (BRB) failure in squirrel cage induction motors (SCIMs) and proposes a new fault diagnosis scheme for fast and instantaneous detection of BRB failures at no-load condition. The proposed method in this paper monitors the condition of a lightly loaded SCIM in order to detect the rotor bar failure process at its early stages. Absolute value of Approximation coefficient of wavelet decomposition at seventh level is considered as a new fault indicator. This indicator is immune to load conditions and can reduce the calculation time without loss of accuracy and resolution. Finite element analysis, MATLAB/SIMULINK and Differential Evolution (DE) optimization algorithm are employed in this work to model the breakage of the bar at the instant of happening. The effectiveness and accuracy of the proposed method for instantaneous detection of rotor bar breakage at no-load condition are evaluated using simulations.