Identification of loosen bolt of transmission tower based on wavelet packet analysis and neural net

The method proposed in this paper identifies the loosen bolt of a transmission tower using the wavelet packet analysis and the neural net. First, the strain signal is measured. Second, the signal is decomposed by the wavelet packet (WP) analysis to extract the WP component energy, which is acted as the subset of the neural net. Finally, the loosen bolt is identified by the neural net. A three-level simple transmission tower is used as the experimental model, and the result show that the method can be effectively used to identify the loosen bolt of transmission tower.