摘要
针对由非线性、非平稳、多状态相依和动态分布等特点导致的交流接触器状态难以准确划分问题,提出了一种基于TSS-DTW的交流接触器状态划分方法。探索采用斜率增量分段线性表示方法提取分割点,完成状态序列的时序分割。在此基础上,利用相邻子序列动态时间弯曲相似性度量形成关键分割点DTW相似性集合,并进行关键分割点鉴别遴选,实现接触器状态的准确划分。实例分析表明,所提方法可实现交流接触器状态划分,且划分界限清晰、类间差异性显著,状态划分精度优势明显。
To address the problem of difficulties in accurately classifying the state of an AC contactor caused by the characteristics of nonlinearity,nonsmoothness,multi-state dependence and dynamic distribution,a state classification method based on TSS-DTW for AC contactors was proposed.It was explored to use the slope incremental piecewise linear representation method for the extraction of segmentation points and complete the timeseries segmentation of the state sequence.On this basis,by using the dynamic time warping similarity metric of neighboring subsequences,the DTW similarity set of key segmentation points was formed,and the key segmentation points were identified and selected,so as to realize the accurate classification of AC contactor states eventually.It is proved that the as-proposed method can realize the state classification of AC contactors,the classification boundary is clear,and the difference between classes is significant,thus the superiority of the state classification accuracy is obvious.
作者
刘洋
曹云东
刘树鑫
LIU Yang;CAO Yundong;LIU Shuxin(School of Electrical Engineering,Shenyang University of Technology,Shenyang 110870,Liaoning,China)
出处
《沈阳工业大学学报》
CAS
北大核心
2024年第6期760-765,共6页
Journal of Shenyang University of Technology
基金
国家自然科学基金项目(51977132)
辽宁省科技重大专项(2020JH1/10100012)。
