Looseness Monitoring of Transmission Tower Based on MIMU

Transmission towers and critical components of electrical grids require health monitoring to ensure structural integrity and reliability. This article presents a novel approach for monitoring the looseness of transmission towers using miniature inertial measurement unit (MIMU) signal analysis based on clustering dynamic mode decomposition (DMD). The proposed approach extracts a key looseness feature, the degree of freedom mode shape (DFMS), by clustering the same modes across different signal blocks. Utilizing the characteristics of transmission tower looseness, we calculate the shape distance of DFMS between different tower floors, enabling the accurate identification of the loose floor. The monitoring process generates vast volumes of multidimensional data. To address this challenge, we introduce an adaptive block compression sensing (ABCS) framework. The success of the ABCS technique relies heavily on a suitable sparse dictionary and sampling strategy. In this article, we introduce an innovative adaptive double sparse K singular value decomposition (ADS-KSVD) dictionary model and an adaptive block observation matrix to efficiently enhance the effect of signal compression and reconstruction. The proposed approach is validated through experiments on a tower modal and an actual tower, yielding satisfactory results in the identification of tower looseness. This article provides an effective and practical solution to ensure the structural safety and reliability of critical infrastructure such as transmission towers.