Research on loose bolt localization technology for transmission towers

Research on loose bolt fault in transmission towers is difficult to detect, and the localization of the loose bolts has always been a challenge in the field of structural health monitoring. This article investigates the relationship between the modal frequencies and modal shapes at different levels and finds that when bolts in transmission towers become loose, the order of modal frequency changes is consistent with the order of the mode shape at the location of the loosening. Based on this, a new method for locating loose bolts in transmission towers is proposed. In the identification of mode parameters, an improved variational mode decomposition is used to extract the mode components at different orders. A modal frequency identification method based on feature set signal reconstruction is proposed, which is based on the modal correlation between the signal components from different sensors, effectively preserving and extracting the structural modal characteristics from the vibration response. At the same time, the identification of modal shapes in transmission towers is achieved through feature-level fusion based on adaptive multi-sequence signals, greatly reducing identification errors. The effectiveness of this method is validated in the finite element analysis of transmission towers, and its superiority is demonstrated through comparison with more advanced methods. The results of a 110 kV transmission tower bolt loosening test also show that this method can accurately identify the area where the bolts are loose, with high identification accuracy, providing a new approach for the localization of loose bolts in transmission towers.