论文部分内容阅读
针对相位敏感光时域反射计(φ-OTDR)分布式光纤扰动传感系统对扰动事件进行有效判别和识别的问题,提出一种基于支持向量机(SVM)的扰动判别和扰动模式识别的方法。通过提取信号时域和频域的平均值、方差、均方差以及信号功率特征,利用二叉树结构建立基于SVM算法的分类器,对扰动进行判别并对扰动模式进行识别。根据传感信号的特征,通过分类器I在对有无扰动信号进行判别的基础上,进一步对有扰动信号利用分类器对扰动事件的模式进行识别。通过实验对所提出的方法进行验证,对600组实验数据进行扰动判别和模式识别,正确的扰动判别率在96%以上,漏报率和误报率在4%以下;正确的模式识别率均在94%以上。
Aiming at the problem that the distributed optical fiber disturbance sensor system of phase-sensitive optical time-domain reflectometer (φ-OTDR) effectively discriminates and identifies disturbance events, a method of disturbance identification and disturbance pattern recognition based on Support Vector Machine (SVM) is proposed. . By extracting average, variance, mean square error and signal power characteristics of signal in time domain and frequency domain, a binary classifier based on SVM algorithm was used to classify the disturbances and identify the disturbance patterns. According to the characteristics of the sensing signal, the classifier I is used to identify the perturbation signal based on the discrimination of the perturbed signal and the classifier. The proposed method is validated by experiments, and perturbation discrimination and pattern recognition are performed on 600 groups of experimental data. The correct disturbance rejection rate is over 96%, and the false negative rate and false positive rate are below 4%. The correct pattern recognition rate Above 94%.