Pattern Recognition Using Relevant Vector Machine in Optical Fiber Vibration Sensing System

Wang, Yu; Wang, Pengfei; Ding, Kai; Li, Hao; Zhang, Jianguo; Liu, Xin; Bai, Qing; Wang, Dong; Jin, Baoquan

doi:10.1109/access.2018.2889699

Cited by 55 publications

(21 citation statements)

References 26 publications

(27 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It can be seen that our proposed method achieves the highest accuracy, and the time of feature extraction is slightly higher than wavelet energy spectrum. This is mainly because in our experiment, we use wavelet energy spectrum method to decompose the signal into two layers instead of six layers in ref [16] to obtain the highest recognition rate.…”

Section: Experimental Results and Discussion A Determination Ofmentioning

confidence: 99%

“…Through replacing the soft-max classifier in CNN with a nonlinear SVM classifier or a linear SVM classifier, the identification rate exceeded 93.3%. In 2019, Wang et al [16] used relevant vector machine (RVM) based on a 7-dimensional feature vector extracted by wavelet energy spectrum analysis to identify three disturbance events (walking through the fiber, striking on the fiber, and jogging along the fiber). A classification macro-accuracy of 88.60% was finally obtained through 10fold cross validation.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Multi-Class Disturbance Events Recognition Based on EMD and XGBoost in φ-OTDR

et al. 2020

View full text Add to dashboard Cite

A novel pattern recognition method based on Empirical Mode Decomposition (EMD) and extreme gradient boosting (XGBoost) is proposed to recognize the disturbance events in phase sensitive optical time-domain reflectometer (ϕ-OTDR) to reduce nuisance alarm rate (NAR) and improve real-time performance in this paper. Eleven typical eigenvectors are extracted from components obtained by EMD of the disturbance signals and XGBoost is selected as a classifier to identify different type of disturbance signals. Five kinds of disturbance events, including watering, knocking, climbing, pressing and false disturbance event, can be identified, effectively. Experimental results show that NAR is 4.10% and identification time is 0.093 s. The recognition accuracy for the five patterns is 97.96%, 95.90%, 91.10%, 94.84% and 99.69%, respectively. The effectiveness of the proposed method is evaluated by using confusion matrix and decision boundary visualization. Experimental results demonstrate that our proposed pattern recognition method based on XGBoost has better performance in recognition rate and recognition time than other commonly used methods, such as support vector machine (SVM), Gradient Boosting Decision Tree (GBDT), Random Forest (RF) and Adaptive Boosting (Adaboost). INDEX TERMS Phase-sensitive optical time-domain reflectometer (ϕ-OTDR), extreme gradient boosting (XGBoost), nuisance alarm rate (NAR), empirical mode decomposition (EMD), pattern recognition, decision boundary visualization.

show abstract

Section: Experimental Results and Discussion A Determination Ofmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Multi-Class Disturbance Events Recognition Based on EMD and XGBoost in φ-OTDR

et al. 2020

View full text Add to dashboard Cite

show abstract

“…This procedure is called the one-againstone approach. Recently, Wang et al report another RVM-based DOVS pattern classification system [96]. However, the structure of multi-class RVM classifier was not described clearly.…”

Section: B Classification Algorithmsmentioning

confidence: 99%

Data-Driven Distributed Optical Vibration Sensors: A Review

et al. 2020

View full text Add to dashboard Cite

Distributed optical vibration sensors (DOVS)have attracted much attention recently since it can be used to monitor mechanical vibrations or acoustic waves with long reach and high sensitivity. Phase-sensitive optical time domain reflectometry ( -OTDR) is one of the most commonly used DOVS schemes. For -OTDR, the whole length of fiber under test (FUT) works as the sensing instrument and continuously generates sensing data during measurement. Researchers have made great efforts to try to extract external intrusions from the redundant data. High signal-to-noise ratio (SNR) is necessary in order to accurately locate and identify external intrusions in -OTDR systems. Improvement in SNR is normally limited by the properties of light source, photodetector and FUT. But this limitation can also be overcome by post-processing of the received optical signals. In this context, detailed methodologies of SNR enhancement post-processing algorithms in -OTDR systems have been described in this paper. Furthermore, after successfully locating the external vibrations, it is also important to identify the types of source of the vibrations. Pattern classification is a powerful tool in recognizing the intrusion types from the vibration signals in practical applications. Recent reports of -OTDR systems employed with pattern classification algorithms are subsequently reviewed and discussed. This thorough review will provide a design pathway for improving the performance of -OTDR while maintaining the cost of the system as no additional hardware is required.

show abstract

“…As one typical distributed optical fiber vibration sensing system, a phase-sensitive optical time domain reflectometer (Φ-OTDR) can achieve simultaneously multi-point vibration detection and location. Because of the benefits of its high sensitivity, good spatial resolution and long detecting distance [4,5], Φ-OTDR has broad application prospects in long-distance oil and gas pipelines [6], border security intrusion detection and smart grids [7,8]. The current research hotspot of Φ-OTDR is the improvement of its sensing performance, such as a sensing distance up to even hundreds of kilometers [9,10], and a spatial resolution of a sub-meter magnitude [11], which leads to a huge amount of sensing data and a growing demand for computation resources.…”

Section: Introductionmentioning

confidence: 99%

Co-Processing Parallel Computation for Distributed Optical Fiber Vibration Sensing

Wang

Jin

et al. 2020

Applied Sciences

Self Cite

View full text Add to dashboard Cite

Rapid data processing is crucial for distributed optical fiber vibration sensing systems based on a phase-sensitive optical time domain reflectometer (Φ-OTDR) due to the huge amount of continuously refreshed sensing data. The vibration sensing principle is analyzed to study the data flow of Rayleigh backscattered light among the different processing units. A field-programmable gate array (FPGA) is first chosen to synchronously implement pulse modulation, data acquisition and transmission in parallel. Due to the parallelism characteristics of numerous independent algorithm kernels, graphics processing units (GPU) can be used to execute the same computation instruction by the allocation of multiple threads. As a conventional data processing method for the sensing system, a differential accumulation algorithm using co-processing parallel computation is verified with a time of 1.6 μs spent of the GPU, which is 21,250 times faster than a central processing unit (CPU) for a 2020 m length of optical fiber. Moreover, the cooperation processes of the CPU and GPU are realized for the spectrum analysis, which could shorten substantially the time of fast Fourier transform analysis processing. The combination of FPGA, CPU and GPU can largely enhance the capacity of data acquisition and processing, and improve the real-time performance of distributed optical fiber vibration sensing systems.

show abstract

Pattern Recognition Using Relevant Vector Machine in Optical Fiber Vibration Sensing System

Cited by 55 publications

References 26 publications

Multi-Class Disturbance Events Recognition Based on EMD and XGBoost in φ-OTDR

Multi-Class Disturbance Events Recognition Based on EMD and XGBoost in φ-OTDR

Data-Driven Distributed Optical Vibration Sensors: A Review

Co-Processing Parallel Computation for Distributed Optical Fiber Vibration Sensing

Contact Info

Product

Resources

About