Abstract
The pressure signal characteristics in the low-pressure gas pipeline inside building are covered up by complex noise, and it is difficult to collect leakage samples. This paper established a complete low-pressure pipeline system leakage detection method based on semi-supervised anomaly identification theory. The proposed method includes signal processing based on combinatorial strategies, semi-supervised model based on deep sequence-to-sequence autoencoder, and anomaly diagnosis based on two-dimensional indicators. Bahdanau's attention mechanism and the long and short-term memory neural network were jointly applied to the semi-supervised model. By improving the Grubbs' test, the designed anomaly diagnosis strategy allows data-driven and manual intervention to make joint decisions. Importantly, we selected actual low-pressure gas pipelines in civil buildings to collect pressure signals to verify the practicability of the proposed method. The proposed method can solve the interference caused by complex noise without leakage samples. Based on actual data, it can finally reach 85.3% leak detection accuracy. In addition, we compared three typical pipeline leakage detection models and three abnormal diagnosis indicators. The results show that the proposed method has obvious advantages in the accuracy of low-pressure gas pipeline leakage detection and the dependence of leakage samples.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.