Bearing diagnosis based on Mahalanobis–Taguchi–Gram–Schmidt method

摘要: Abstract A methodology is developed for defect type identification in rolling element bearings using the integrated Mahalanobis–Taguchi–Gram–Schmidt (MTGS) method. Vibration data recorded from with seeded defects on outer race, inner race and balls are processed time, frequency, time–frequency domains. Eleven damage parameters (RMS, Peak, Crest Factor, Kurtosis time domain, amplitude of ball frequencies FFT spectrum HFRT frequency domain peak HHT domain) computed. Using MTGS, these (DIPs) fused into a single DIP, Mahalanobis distance (MD), gain values presence all DIPs calculated. The value used to identify usefulness DIP positive again MD by Gram–Schmidt Orthogonalization process (GSP) order calculate Vectors (GSVs). Among remaining DIPs, sign GSVs checked classify probable defect. approach uses MTGS method combining conjunction GSV classifies Defect Occurrence Index (DOI) proposed rank probability existence bearing (ball defect/inner defect/outer defect/other anomalies). successfully validated vibration different machine, shape/configuration also applied acquired accelerated life test bearings, which established applicability naturally induced progressed It observed that identifies correct useful identifying initiation has potential implementation real environment.