Application of SLAM in bolt tightening monitoring process based on intelligent data algorithm analysis

Abstract

In this paper, firstly, according to the requirements of static monitoring of bolt tightening quality, the bolt tightening process can be divided into four stages: snug, elastic deformation, yielding and plastic deformation, and the SLAM technology is applied to bolt monitoring modeling and building engineering modeling. Next, the pre-tightening force for the bolt is measured using the waveform expansion method, and the bolt tightening index can be determined. Then, all the collected bolt data are randomly combined into two categories of loosening and tightening to form a training data set, and the improved intelligent data algorithm is used to train and predict them. The results show a linear correlation between tightening degree x |τ ɛ| and preload force change ΔF under 0.31kN and 0.22kN loading gradients, and the linear correlation coefficient of the fitted curve is greater than 0.9215. The results of the data analysis verify the correctness of the theoretical analysis.