摘要
为实现低压万能式断路器储能操作机构的机械寿命预测,提高其运行可靠性,提出了一种基于机电信号多特征融合的寿命预测方法。首先对储能操作机构动作过程进行分析,从电流、振动信号多角度反映储能操作机构的退化状态信息,结合信号特点提取时域和频域的多个退化特征,并获取相关性高的关键退化特征。在此基础上,基于关键退化特征的统计分析,实现对退化过程的有效判定。最后,对关键特征进行融合,得到综合健康指标,在进入退化期后作为一元回归指数模型的输入,以实现剩余机械寿命的定量预测。3台试品测试结果表明:寿命预测平均绝对误差在35次以下,平均相对误差在11.7%以下,该方法有效提升了寿命预测的精度并具有工程实用性。
In order to realize the prediction of mechanical life of energy storage operating mechanism of low-voltage conventional circuit breaker and improve its operation reliability,a life prediction method based on multi-feature fusion of electromechanical signals is proposed.Firstly,the action process of energy storage operating mechanism is analyzed,the degradation state information of energy storage operating mechanism is reflected from multiple angles of current and vibration signals,multiple degradation features in time domain and frequency domain are extracted combined with the signal characteristics,and the key degradation features with high correlation are obtained.On this basis,based on the statistical analysis of key degradation features,the effective determination of degradation process is realized.Finally,the key features are fused to obtain a comprehensive health index,which is used as the input of the unitary regression exponential model after entering into the degradation period,so as to realize the quantitative prediction of the remaining mechanical life.The test results of three samples show that the mean absolute error of life prediction is less than 35 times and the mean relative error is less than 11.7%.This method not only can be utilized to effectively improve the accuracy of life prediction but also has engineering practicability.
作者
孙曙光
唐尧
王景芹
温志涛
高辉
SUN Shuguang;TANG Yao;WANG Jingqin;WEN Zhitao;GAO Hui(School of Artificial Intelligence,Hebei University of Technology,Tianjin 300130,China;State Key Laboratory of Reliability and Intelligence of Electrical Equipment,Hebei University of Technology,Tianjin 300130,China;Tianjin Benefo Electric Co.,Ltd.,Tianjin 300385,China)
出处
《高电压技术》
EI
CAS
CSCD
北大核心
2022年第11期4455-4468,共14页
High Voltage Engineering
基金
河北省自然科学基金(E2021202136)。
关键词
万能式断路器
剩余寿命
机电信号
特征提取
主成分分析
一元线性回归
conventional circuit breaker
remaining useful life
electromechanical signal
feature extraction
principal component analysis
unary linear regression
