US 11,959,978 B2
Method of detecting a rotor bar fault and a method of estimating an additional operating expenditure due to one or more mechanical anomalies in an electrical machine
Subash Chandar Athikessavan, Singapore (SG); Sanjib Kumar Panda, Singapore (SG); and Shiva Shankaranarayanan Muthuraj, Singapore (SG)
Assigned to Sembcorp Industries Ltd, Singapore (SG); and National University of Singapore, Singapore (SG)
Appl. No. 17/632,760
Filed by Sembcorp Industries Ltd, Singapore (SG); and National University of Singapore, Singapore (SG)
PCT Filed Aug. 4, 2020, PCT No. PCT/SG2020/050452
§ 371(c)(1), (2) Date Feb. 3, 2022,
PCT Pub. No. WO2021/025620, PCT Pub. Date Feb. 11, 2021.
Claims priority of application No. PCT/SG2019/050389 (WO), filed on Aug. 5, 2019.
Prior Publication US 2022/0276307 A1, Sep. 1, 2022
Int. Cl. G01R 31/72 (2020.01); G01R 31/34 (2020.01); G01R 31/52 (2020.01)
CPC G01R 31/72 (2020.01) [G01R 31/343 (2013.01); G01R 31/346 (2013.01); G01R 31/52 (2020.01)] 25 Claims
OG exemplary drawing
 
1. A method of detecting a rotor bar fault of an electrical machine, said method comprising,
acquiring a set of online signals from the electrical machine over a period, said set of online signals comprising at least one vibration signal, and a summed signal of magnetic flux signals, each magnetic flux signal obtained from a respective flux sensor positioned on an external surface of the electrical machine,
wherein the summed signal of magnetic flux signals is obtained by acquiring and summing a set of p flux signals to obtain a total flux signal, wherein the set of p flux signals is acquired from p flux sensors positioned along a circumference of a cylindrical frame of the electrical machine, such that any two adjacent flux sensors along the circumference of the cylindrical frame have an angular separation of 360/p degrees with respect to the longitudinal axis of the cylindrical frame, wherein p represents the total number of poles of the electrical machine;
extracting an online anomaly indicator value from the set of online signals;
comparing the online anomaly indicator value with a baseline anomaly indicator value of a corresponding loading condition; and
determining presence of a broken or cracked rotor when the online anomaly indicator value deviates from the baseline anomaly indicator value by a threshold.