摘要
针对低压断路器的机械特性,采用小波分解法对其振动信号进行分析。根据电动操作机构及低压断路器合闸动作的时序关系,以驱动电机电流信号作为时间标识,有效地提取了合闸振动信号。提出小波包能量谱分析低压断路器合闸同期性研究,在小波包对振动合闸信号细节分解基础上,采用小波包重构提取合闸振动主频带信号特征,由此构造合闸同期性状态特征矢量,并应用BP神经网络建立三相合闸不同期故障的识别模型。在断路器基座横梁安装单个加速度传感器,实验模拟了DW15—1600低压断路器的四种同期性状态振动信号,仿真结果表明,本文提出的振动信号小波包能量谱与神经网络相结合的方法,可有效地分析低压断路器合闸同期性。
Wavelet decomposition method is used to analysis the low voltage circuit breaker mechanical properties with its vibration signals. According to the electric operating mechanism of low voltage circuit breaker and its closing action sequence relations, driving motor current signal as a time stamp is applied to effectively extract closing vibration signal. A novel low voltage circuit breaker closing synchronous research is proposed with wavelet energy spectrum analysis in this paper. Based on refine decomposition to closing vibration signal and feature extraction from its main frequency band with wavelet packet reconstruction, the feature vector of closing synchronous is constructed. Three phases closing asynchronous fault identification model is established by back propagation neural network with above feature vector. The vibration signals of a DW15-1600 low-voltage circuit breaker under four specific closing synchronous status simultaneities are recorded from a single acceleration sensor mounted on a cross beam of breaker base. The simulation results show that the combination method of wavelet packet energy spectrum and neural network can effectively analysis the closing synchronism of a low voltage circuit breaker.
出处
《电工技术学报》
EI
CSCD
北大核心
2013年第6期81-85,共5页
Transactions of China Electrotechnical Society
基金
福建省高校产学合作重大科技资助项目(2011H6013)
关键词
低压断路器
振动分析
小波分解
神经网络
合闸同期性
故障识别
Low voltage circuit breaker
vibration analysis
wavelet decomposition
neural networks
closing synchronous
fault identification
