Application of clustering analysis in distributed fiber optical vibration sensing systems

To solve the problem that it is difficult to position the perturbation point and judge the perturbation time under multiple simultaneous perturbations in a distributed optical fiber vibration sensing system, signal processing method was carried out and clustering analysis technology was applied. Firstly, hotspot map gathered by the system was processed above a threshold, and then the target objects were extracted. According to the results of clustering analysis, perturbation points are positioned and perturbation time is confirmed. The results show that the method makes an accurate and reasonable judgment for positioning the perturbation behavior and perturbation duration and the noise interference is eliminated at the same time. This is one step of event classification. The study provides the basis for the final system alarm strategy.