基于变分模态分解和鲁棒性独立成分分析的内燃机缸盖振动信号分离

针对缸盖振动信号中燃烧信号和活塞敲击信号在时频域混叠严重难以分离的问题，用基于VMD和RobustICA的方法分离各独立源信号。通过试验测量内燃机单通道缸盖振动信号，首先对测得的信号进行消除趋势项及滑动平均等预处理，用VMD算法对预处理后的信号进行分解；然后用RobustICA算法提取独立成分，并用组合模态函数法对时域和频域相似性较高的分量成分进行组合；最后结合频谱分析、连续小波变换、相干函数法及倒拖试验对分离得到的结果进行识别验证。研究结果表明：在不同的试验工况下，该方法可以有效地从缸盖振动信号中分离出燃烧信号和活塞敲击信号。

Cylinder Head Vibration Signals Separation of Internal Combustion Engines Based on VMD and RobustICA

In order to solve the problems that the combustion signals and the piston slap signals in cylinder head vibration signals were difficult to separate in the time-frequency domain, the VMD and RobustICA methods were utilized to separate the independent source signals. The single channel cylinder head vibration signals of an internal combustion engine were measured by tests. Firstly, the measured signals were carried out the pretreatments, such as eliminations of trend items and move average. The VMD algorithm was employed to decompose the vibration signals. Then the RobustICA algorithm was applied to extract the independent components. The CMF method was adopted to combine the obtained independent components which were similar in time domain and frequency domain. Finally, the spectrum analysis, the continuous wavelet transform, the coherence function method and the drag test were utilized to further identify the separation results. The research results show that this method may effectively separate the combustion signals and the piston slap signals from cylinder head vibration signals under different test conditions.