Oil Pipeline Safety Monitoring Method based on Vibration Signal Analysis and Recognition

An analytically tractable model is presented to describe oil and gas pipeline safety monitoring system. The basic idea is that it uses fiber sensor to collect signals produced by soil vibration around pipelines, and then focuses on intelligent processing and smart recognition of soil vibration signals. Finally, we implement the developed model and it is practically used in Jiangsu, China. Experiment results show that the system can fully satisfy the real time requirement. Further more, the alarm rate is higher than 98% and the recognition rate is 95.3% for five different kinds of human activities (ramming, picking, drilling, steel pipe knocking, forklift working), much better than other results reported yet.