摘要
水轮机组状态监测数据由于受到传感器、测量条件、环境等因素影响会造成数据失真,产生众多奇异点,对正确分析机组运行状态十分不利。本文提出通过经验模态分解与小波变换相结合的方法来分析水轮机故障信号的奇异性,该方法将原始信号经验模态分解后,利用小波变换检测出信号中的奇异点,并将剔除奇异点的信号重构,通过重构信号对机组进行分析。实例仿真表明,与直接对原信号进行小波分析相比,该方法提取的奇异性特征明显并能准确重构,在采用通用传感器信号准确描述机组状态和正确认识机组故障上有较好的应用价值。
State-monitoring data of a hydraulic generator under the factors of sensor,measurement condition and environment are often distorted and singular data are generated.Wavelet transform is an effective way to correctly detect signal singularity.This paper develops a method combining mode decomposition and wavelet transform.In this method,singular signal is detected by wavelet transform,then the singular points are removed and new data are reconstructed using the signal details and certain approximate coefficients.Simulation results show a good agreement of the reconstructed signal with the original one.The proposed EMD-wavelet analysis on monitoring data has a practical value and it is vital to fault recognition and description of hydraulic generator state.
出处
《水力发电学报》
EI
CSCD
北大核心
2012年第2期245-250,共6页
Journal of Hydroelectric Engineering
基金
国家自然科学基金项目(50779056)
关键词
水电机组
经验模态分解
小波变换
奇异性
信号处理
故障诊断
hydraulic generator
empirical mode decomposition
wavelet transform
singularity
signal process
fault diagnosis
