A. GMM-Based intrusion detection for perimeter security system

As the exploitation of shale gas has increased in recent years, the safety of perimeter protection for shale gas wells has become increasingly important. Today, there are a number of problems with algorithms that build models based on foreground signals for intrusion detection. The biggest problem is that the use of foreground modeling requires a large number of intrusion signals, but often in the actual environment, the intrusion signal is relatively small, the signal will be in a normal state without interference for a long time, and the signal waveform generated by different intrusion methods is also different. This can result in low intrusion detection rates and scattered models. Therefore, in this paper, based on the background modeling method, the perimeter detection of wellsite intrusion detection is studied. By collecting data in the field, preprocessing the data and designing the algorithm to verify the feasibility of the method. Experiments show that the proposed background intrusion detection rate based on background modeling can reach 90.56%, which greatly meets the basic requirements for intrusion detection.