Fault Diagnosis of Loose Anchor Bolt of Coal Mine Ventilator Based on EMI

Aiming at the difficulty of identifying bolt loosening, a new method for diagnosing anchor bolt loosening based on piezoelectric impedance technology is proposed. Firstly, a mathematical theoretical model of coupling PZT and bolt was constructed to explain the causes of the deviation of the admittance curve, and the piezoelectric impedance meter was used to collect the admittance values of bolts under different states, and the resonant peak characteristics of the admittance curve were obtained to verify the reliability of the constructed model. Secondly, RMSD damage evaluation index was used to process admittance data to judge the number of bolt loosening. Finally, the PSO-BP neural network is constructed and trained to realize the recognition of bolt loosening conditions. The experimental results show that the piezoelectric impedance technique can effectively identify the loosening condition of bolts. After training, the PSO-BP neural network can quickly and accurately identify various bolt loosening conditions.