Assessment of Pressure Vessel Based on Variational Bayesian

Haonan Zhang,Chun Yin,Yuhua Cheng,Xuegang Huang,Kai Chen,Anhua Shi,Xuan Gou
DOI: https://doi.org/10.1109/i2mtc43012.2020.9128445
2020-01-01
Abstract:Pressure vessels play an important role in daily life. However, pressure vessels exposed to high temperature and pressure for a long time are prone to some damages, such as micro-cracks, corrosion-fatigue crack, etc. In order to identify and assess the damage of pressure vessel, a damage assessment framework based on pulsed eddy current data is proposed in this paper. Firstly, a damage reconstruction model is propsoed according to the infrared image sequence of pressure vessel, and the damage images of different layers are obtained by using the Variational Bayesian. Then, image smoothing and segmentation method is used to eliminate the reconstructed noise and data acquisition noise. Finally, image fusion method is used to generate images describing all damage of the pressure vessel. In the experiment, a part of the pressure vessel is used for damage assessment, and the image of each step shows the effectiveness of the proposed method.
What problem does this paper attempt to address?