RVFL-Based Optical Fiber Intrusion Signal Recognition With Multi-Level Wavelet Decomposition as Feature

Abstract

The optical fiber pre-warning system (OFPS) has been gradually considered as one of the important means for pipeline safety monitoring. Intrusion signal types are correctly identified which could reduce the cost of troubleshooting and maintenance of the pipeline. Most of the previous feature extraction methods in OFPS are usually quested from the view of time domain. However, in some cases, there is no distinguishing feature in the time domain. In the paper, firstly, the intrusion signal features of the running, digging, and pick mattock are extracted in the frequency domain by multi-level wavelet decomposition, that is, the intrusion signals are decomposed into five bands. Secondly, the average energy ratio of different frequency bands is obtained, which is considered as the feature of each intrusion type. Finally, the feature samples are sent into the random vector functional-link (RVFL) network for training to complete the classification and identification of the signals. Experimental results show that the algorithm can correctly distinguish the different intrusion signals and achieve higher recognition rate.

Random vector functional link network: recent developments, applications, and future directions

“…In the optical fiber pre-warning system (OFPS), most of the feature extraction methods are quested from the view of the time domain. To address this issue, using multi-level wavelet decomposition, Wang et al [85] extract intrusion signal features of the running, digging, and pick mattock in the frequency domain and then for considering the feature of each intrusion type, the average energy ratio of different frequency bands is obtained. Finally, the RVFL model is trained for the classification and identification of the signal.…”

“…Newton-Armijo stepsize method Classification problem Sahani and Dash [99] Class-specific weighted RVFL (CSWRVFL) Tanh Closed form (LS) Power quality disturbances Fan et al [54] Random subspace fisher linear discriminant (FLD) based RVFL -Closed form (LS) Image steganalysis Wang et al [85] Standard RVFL -Closed form (LS) Optical fiber pre-warning system Pratama et al [84] Parsimonious RVFL (pRVFL) -FWGRLS method Data stream Henriquez and Ruz [83] Neural networks with random weights (NNRW) Sigmoid Closed form (LS) Regression and classification problem Vuković et al [31] Orthogonal polynomial expanded RVFL (OPE-RVFL) Tansig, logsig, tribas Closed form (ridge regression/ Moore-Penrose pseudoinverse)…”

An Improved High-Intelligence Method of Gas and Oil Pipeline Prewarning System in Real Soil Environment

“…is process of extracting signal features takes a lot of time. Second, the NN in [29][30][31] used a single NN, which is hard to effectively learn the signal characteristics collected by the DAS. Besides, the existing experiments in [32][33][34][35] only utilized the data being generated very close to the sensed area (tens or hundreds of meters), which had a small sensing area or were at a location with a clean background environment.…”

Research on Intelligent Optical-Fiber Pre-Warning System for Long-Distance Pipeline Safety

“…。 中国的油气输送已形成了贯穿全国，联通海外的传输 网。虽然通过管道运输有效地提升了油气传输的效率， 但一旦管道遭到破坏， 极易发生供应中断， 环境污染， 爆炸等事故，造成巨大的经济损失、自然灾害和人员 伤害 [2] 。管道安全预警系统主要用于预警对管道安全 造成威胁的入侵事件。除管道腐蚀、山体滑坡等自然 灾害之外，机械、人工挖掘，打孔等第三方破坏事件 是造成管道泄漏的主要因素 [3] [4][5] 。 因此在实际应用中常使用性价比更高的光 纤传感系统来监测破坏管道安全的入侵事件 [4] 。根据 光学原理的不同，适用于管道安全监控领域的光纤传 感方法主要分为散射法和干涉法。从实际应用出发， -OTDR 型分布式光纤传感系统定位精度较高且易 于部署，监测距离较长且对外界微弱的振动信号也有 较高的灵敏度。系统可以对光纤沿线多处发生的振动 信号实现同时检测，安全预警技术性价比最高 [6][7][8] 。因 此本文采用-OTDR 分布式光纤传感系统采集沿管 道周围的土壤振动信号 [9] 。 分布式光纤传感系统较高的定位识别精度和长 距离的监测应用使得系统每毫秒接收的数据量较大， 对管道破坏性事件的识别也极易受到周围复杂环境 的汽车、火车等无规律过车振动信号的影响，极易造 成较多的误报。因此需要一种简单和易于推广到各 种应用环境的过车事件识别算法来处理传感系统采 集到的大量的数据，并能够在复杂环境中快速识别 和定位过车事件。不同的算法对过车事件识别的准 确性存在着明显的差异 [10] 。早期的研究主要通过软 硬阈值 [11][12][13] ，利用基频周期模型进行决策树分类 [14][15] 及支持向量机(support vector machines, SVM) [16] 的 方式实现过车信号的识别。但由于实际环境中过车信 号及其他入侵信号具有未知性和多样性，这些方法对 过车信号识别效果较差。直到近期国内外的一些研究 才开始将神经网络和过车信号的识别结合起来。但目 前这些研究仅使用一种神经网络 [17][18][19][20][21] ，并不能有效地 学习光纤传感系统采集到的过车信号特征。此外， -OTDR 光纤传感系统检测到的信号强度会随着空 间距离的增加呈指数下降 [22] ，然而现有的文献报道中 所做的验证实验大多都在距离光纤数十或数百米的 位置生成数据 [19][20][21]23] ，背景环境较为干净 [20][21]23] The vehicle pa…”