Detecting Abrupt Changes in a Noisy van der Pol Type Oscillator
Detecting Abrupt Changes in a Noisy van der Pol Type Oscillator
复制标题
DOI：
10.1007/s12591-015-0252-z
复制
发表时间：
2015-06
影响因子：
1
通讯作者：
Salam Nema;P. Kowalczyk
中科院分区：
文献类型：
--
作者：
Salam Nema;P. Kowalczyk
动力系统产生的许多信号可能会发生突变，如跳跃或急剧变化。在从探索性数据分析到诊断的许多应用中，检测这些变化点是至关重要的。本文讨论了作为非线性电阻电路模型的噪声范德波尔振荡器突变的检测。该方法将小波分析与信息熵相结合，提取出突变信号对应的频率。我们还研究了噪声强度对模型系统中变化点检测的影响。利用不同模型参数生成的模拟数据对性能进行了评价。
Many signals produced by dynamical systems may undergo abrupt changes such as a jump or a sharp change. Detecting such change points is of primary importance in many applications ranging from exploratory data analysis to diagnosis. This paper addresses the detection of abrupt changes in a noisy van der Pol oscillator as a model of an electrical circuit with nonlinear resistance. The proposed approach combines wavelet analysis with information entropy in order to extract signal frequencies corresponding to any abrupt changes that occur. We also investigate the influence of noise intensity on detecting change points in the model system. Performance is evaluated on simulated data generated by using different model parameters.