基于RBF网络的油气管道侵入事件识别方法研究

Study on Recognition Method Based on RBF Neural Network for Intrusion Events of Oil and Gas Pipeline

研究一种用于油气管道安全分布式光纤预警系统的侵入事件识别方法。该预警系统基于Mach-Ze-hnder光纤干涉仪原理,沿管道同沟敷设光缆,利用其中的三条单膜光纤构成分布式微振动测试传感器。系统实时地检测管道沿途振动信号,采用基于小波包分析的“能量-状态”法获取振动信号特征,并通过RBF神经网络进行判断是否有侵入事件发生,随后对事发点进行定位。现场实验数据验证了该识别方法的有效性。

A recognition method for intrusion events is studied, which is used in the distributed optical fiber pre-warning system for the safety of oil and gas pipeline. In this pre-warning system, which is based on the principle of MachZehnder optical fiber interferometer, an optical cable is laid along the pipeline in the same ditch and three single mode optical fibers in the optical cable build up the distributed micro-vibrant measuring sensor. The system can judge whether intrusion events have occurred by detecting the vibration signals along the pipeline in real-time, extracting the eigenvectors of the vibration signals by the ＂energy-status＂ method based on wavelet package analysis, and the recognition through the RBF neural network. Subsequently the position of the intrusion events can be located. Finally, the data obtained at oil field prove the effectiveness of this method.