A Detection and Alarm Algorithm Against the Third-Party Excavation

Huijie Guo,Yulong Li,Fan Yang
DOI: https://doi.org/10.1109/icicas48597.2019.00134
2019-01-01
Abstract:It often happens that the third-party excavation damage to underground pipelines in cities. A detection and alarm algorithm to prevent the third-party construction excavation is proposed in order to give early warning for the damage of the third-party excavation. Through the construction of excavation signal database, the time-frequency joint feature vectors of different types of excavation signals are extracted, and the machine learning classification model of excavation signals is constructed with the idea of deep learning to predict and classify common types of the third-party excavation signals, so as to capture and monitor the third-party excavation signals in real time. Once the excavation event is detected to be happening, remote alarm will be carried out to realize the early warning of underground pipeline excavation damage, and then the safe operation of underground pipelines is effectively guaranteed.
What problem does this paper attempt to address?