摘要
为提高柴油机故障诊断的准确性与可靠性,针对含噪声车用柴油机振动信号的瞬时非线性特点,将含噪声柴油机振动信号经验模态分解,去掉主要干扰因素所对应的IMF分量,再将剩余IMF分量进行重构,得到去噪声后柴油机振动信号时间序列;并应用混沌理论,选择合适的时滞τ,对去噪后车用柴油机振动信号时间序列进行相空间重构,并得出了不同嵌入相空间下去噪后车用柴油机振动信号时间序列关联维的变化规律.结果表明:重构的去噪后车用柴油机振动信号能反映柴油机机身振动的真实趋势,是混沌序列,并具有分形特征.在车用柴油机系统中,影响去噪后车用柴油机振动信号的系统内部因素最多可达8个,最小不会小于1个,这为柴油机振动信号的在线故障诊断提供了理论依据.
In order to enhance the accuracy of the fault diagnosis, based on the instantaneous nonlinear characteristics of the vibration signals from the vehicle diesel engine, the vibration signals from the vehicle diesel engine is disposed with EMD method and the IMF components corresponding to main interference factors are eliminated, then the true vibration signal from diesel engine can be gotten by reconstructing the remaining IMF components. And the time-series embedding space about de-noising vehicle engine vibration signals is rebuilt after a suitable value of time lag v has been decided based on the theory of chaos, and their changing regularities of the value of the correction dimension are gotten in different embedding spaces. The results reveal that the signal being reconstructed can reflect the true trend of cylinder block vibration for the diesel engine and there are fractal and chaos features in the de-noising vehicle engine vibration systern, there are eight internal factors at best and one internal factors at least, which provides much available academic basis for on-line fault diagnosis of vibration signals from diesel engine.
出处
《重庆工学院学报（自然科学版）》
2008年第6期6-10,50,共6页
Journal of Chongqing Institute of Technology
基金
湖南大学“985工程”二期--高效低排放发动机先进设计制造技术科技创新平台资助
关键词
去噪
柴油机振动信号
混沌
时间序列
分形
IMF分量
de-noising
Diesel engine vibration signals
chaos
time series
fractal
IMF component
