Study of Fault Detection of Bridge Crane Wheel based on Fourier Transform

The track condition of a bridge crane directly affects the production efficiency and life safety. Due to the limitation of harsh environment, the traditional detection methods include high altitude danger, difficult operation and low efficiency. In this paper, according to the on-line inspection of wheel track wear, the acoustic signal of electric converter is used to collect the sound signal. The method of failure detection based on power spectrum is proposed, and the BIF feature selection combined with Fisher criterion is used to select the best special collection, and the problem of many characteristics is solved. Finally, we use the two classification logic regression to achieve the mathematical modeling between the feature set and the wear volume, and use the H function as the basis of failure judgement. The results show that the failure probability value of the system is more than 0.8 when the wheel wear is serious and close to failure, which is approximately equal to the real value. It can provide a reliable basis for the detection of wheel track.