Research on laser vibration detection and intelligent identification of loose bolts of overhead transmission pole towers

In view of the shortcomings of the traditional transmission line tower bolt loosening detection method, this paper firstly introduces the research background and the existing problems of the existing detection methods, proposes the technology of using long-distance laser vibration measurement device for tower bolt loosening detection and identification, expounds the principle of bolt loosening laser vibration detection, and the basic composition of laser vibration measurement device. In order to improve the recognition accuracy of bolt looseness detection in the tower, the MFCC (Mel-scaleFrequency Cepstral Coefficients) feature extraction algorithm was used to construct the vibration feature model of bolts in different tightness states by combining the GMM-UBM (Gaussian Mixture Model-General Background Model) model, and the clustering characteristics of bolts in different tightness states were analyzed, the clustering characteristics of bolts in different tightness states were obtained, and the identification program was developed. After testing, the accuracy of loose identification of tower bolts ≥ 93%.