一种基于多源感知与机器学习的储氢罐寿命预测方法
By employing multi-source sensing and machine learning methods, the problem of quantifying microscopic damage parameters in hydrogen storage tanks was solved, enabling accurate life prediction under varying operating conditions and improving the health management capabilities of hydrogen storage tanks.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHANGSHA UNIVERSITY
- Filing Date
- 2026-06-22
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies struggle to effectively detect and quantify microscopic damage parameters such as initial crack size and hydrogen embrittlement coefficient in hydrogen storage tanks. Furthermore, the evaluation models exhibit poor generalization ability under varying operating conditions, making it difficult to achieve continuous, automatic tracking and cumulative calculation of damage.
By employing multi-source sensing and machine learning methods, pressure signals, strain signals, and acoustic emission signals are collected simultaneously. The probability distribution of unmeasurable damage parameters is inverted using a physical information neural network. Weighted fusion and Gaussian process regression are then performed through an attention network to construct a cumulative damage trajectory to predict remaining lifespan.
It achieves fast and stable convergence and personalized adaptation of the model under conditions of very few online samples, improving the accuracy and robustness of prediction. It can continuously track the damage state in each hydrogen charge-discharge cycle and output the remaining lifetime with confidence interval.
Smart Images
Figure CN122412931A_ABST