Information fusion with Correlation Coefficient for detecting inter-turn short circuit faults in asynchronous machines

This paper presents a new method giving high efficiency for detecting an inter-turn short-circuit fault in the stator winding of asynchronous machines. For evaluation of the machine state and final decision, the monitoring of the magnetic field variation in the vicinity of an electrical machine is used. The proposed approach is based on the fusion of information extracted from signals delivered by flux sensors placed in different positions around the machine and the calculation of Pearson correlation coefficient. This coefficient allows one to quantify the linear relationship between the signals delivered by two sensors S1 and S2 placed at 180° around the machine in several positions. The proposed approach is non-invasive and relies on the calculation of a correlation coefficient derived from measurements of the external magnetic leakage field for different load working cases. The ability of proposed coefficient to provide useful information about faults is investigated in the paper.