摘要
基于机械振动信号对高压断路器进行故障诊断是一种有效的手段。提出一种基于振动信号的随机森林集成学习模型,进一步提高了故障诊断模型的效率和准确度。首先,获取电磁斥力机构高压断路器两个不同位置的分闸振动信号;然后,利用小波包分解对振动信号进行时、频特性分析,并计算振动信号的各频段归一化能量向量;最后,基于随机森林集成学习模型进行故障诊断和识别。实验结果显示,随机森林集成学习模型具有显著的泛化性能,使用简单,诊断更加快速、准确。
Fault diagnosis of high voltage circuit breakers based on mechanical vibration signals is an effective means.A random forest integrated learning model based on vibration signal was proposed to further improve the efficiency and accuracy of fault diagnosis.Firstly,the brake vibration signals of two different positions of the high-voltage circuit breaker of the electromagnetic repulsive mechanism are obtained,and then the time-frequency characteristics of the vibration signal are analyzed by wavelet packet decomposition,and the normalized energy vector of each frequency band of the vibration signal is calculated,and finally based on the random forest integrated learning model,the fault diagnosis and identification are carried out.The experimental results show that the random forest integrated learning model has significant generalization performance and is simple to use,and the diagnosis is faster and more accurate.
作者
樊浩
苏海博
陈立
史宗谦
李兴文
FAN Hao;SU Haibo;CHEN Li;SHI Zongqian;LI Xingwen(Department of Electrical Engineering,Xi’an Jiaotong University,Xi’an 710049,China;Electric Power Test Research Institute,Guangzhou Power Supply Bureau Co.,Ltd.,Guangzhou 510410,China)
出处
《电器与能效管理技术》
2019年第23期1-5,50,共6页
Electrical & Energy Management Technology
基金
SF6罐式结构真空快速开关状态监测技术研究及应用——罐式SF6绝缘快速真空开关的状态监测评估方法研究及监测样机试制(GZHKJXM20180086)
关键词
小波包归一化能量
随机森林
集成学习
故障诊断
wavelet packet normalization energy
random forest
integrated learning
fault diagnosis