Broken bar fault diagnosis of induction motors using MCSA and neural network

Abstract

Early detection and diagnosis of incipient faults are desirable to ensure an operational effectiveness improved of an induction motors. A novel practical detection and classification method, using motor current signature analysis (MCSA) associated with a neural technique is developed to detect rotor broken bar faults. In this method, only one phase current is used. Following current spectrum study on hundreds of experimental observations, it was established that the mixed eccentricity harmonic f <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ecc_mix</sub> has the largest amplitude around the fundamental, under different loads and state (healthy or defective). However f <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ecc_mix</sub> is related to the slip and the mechanical rotational frequency. It becomes obvious that the detection of the rotor broken bars harmonic is made easy. The amplitude of this harmonic and the slip (detection criterion) are used as the neural network inputs. The last provides reliably, its decision on the state of the machine. Experimental results prove the efficiency of the proposed method.