摘要
以满负载条件下频繁操作的交流接触器为研究对象,提出一种基于数据驱动的电器运行状态监测方法.首先,通过试验平台采集交流接触器历史运行数据,得到状态特征参量,并结合小波变换与主成分分析综合评价,对参量数据进行去噪去奇异值等预处理;其次,针对参量数据存在高维、冗余等负面干扰问题,采用核主成分分析方法进行多信息融合,并基于试验数据进行核参数的优选;最后,将融合信息量输入至隐半马尔可夫模型中,实现智能电器运行状态的监测与识别.以CJX2-8011交流接触器的试验数据为例,验证了所提方法在电器状态监测中的实用性和有效性.
Alternating current（AC）contactor used frequently under full load is taken as a research object,and a method of apparatus operation condition monitoring based on data driven is proposed in this paper.First,historical operation data of AC contactor are collected by test platform and state characteristic parameters are gained.Combined with comprehensive assessment of wavelet transform and principal component analysis,data preprocessing such as denoising and removing outliers is used.Then,aimed at the disadvantage of high-dimensionality and redundancy,kernel principal component analysis（KPCA）is adopted to merge multi-information and kernel parameters are optimized based on test data.Finally,the fused information is input to hidden semi-Markov model（HSMM）,and the operation condition monitoring and recognition of intellectual apparatus are realized.The practicability and validity of the method in apparatus condition monitoring is verified by test data from CJX2-8011 AC contactor.
作者
高筱婷
杨东升
GAO Xiaoting YANG Dongsheng(College of Information Science and Engineering,Northeastern University, Shenyang 110819, China)
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2017年第9期1104-1110,共7页
Journal of Shanghai Jiaotong University
基金
国家自然科学基金(61273029
61433004)
辽宁省自然科学基金(2013020037)
教育部新世纪优秀人才支持计划(NCET-12-0106)
中央高校基本科研基金项目(N130504004)资助
关键词
状态监测
多信息融合
核主成分分析
隐半马尔可夫模型
智能电器
小波变换
condition monitoring
multi-information fusion
kernel principal component analysis（KPCA）
hidden semi-Markov model（HSMM）
intellectual apparatus
wavelet transform
