Angular misalignment is one of the factors that can lead to premature bearing failure. However, existing methods struggle to monitor the degree of misalignment in real time due to the difficulty of installing sensors on such precise equipment. To address this issue, this study proposes an intelligent soft sensor that predicts the degree of angular misalignment based on vibration signals. Specifically, a contrastive attention network is designed to extract latent patterns embedded within the vibration signals. Furthermore, a bearing angular misalignment experimental platform is established. Experimental results demonstrate that the proposed method achieves favorable performance in both measurement accuracy and response speed.

bearing assembly error,angular misalignment,intelligent soft sensor,virtual measurement,deep learning