Gear Fault Signal Modeling Study Based on Probability Box Theory
An Liu, Yi Du, Jia Man Ding · Applied Mechanics and Materials · 2014
Gears typical failure modes and fault diagnosis methods were summarized, and their characteristics and deficiency were contrasted. As almost all method need feature extraction before information fusion, the rich information in original signals were lost in this process. Another difficult problems of information fusion is the the space-time registration. The probability box theory can be a new method to solve the above two problems. The gears fault signal modeling method based on probability box theory were then proposed. Finally the prospects and study directions of this method’s applications in gear box fault diagnosis were proposed.