基于瞬时转速多特征融合的内燃机各缸工作状态识别

为分析内燃机瞬时转速波动过程与各缸工作相位的对应关系,进而识别各缸工作状态,提出了基于硬件计数原理的内燃机瞬时转速精确测试方法,在某自行火炮发动机上测取了发动机正常工作、燃烧不良和失火工况下的瞬时转速信号。对信号进行频谱分析,并利用分频带滤波方法分离出电调信号和各缸做功冲击信号,进而提出5个特征参数并计算各特征参数的相对偏差和偏差极性,最后根据不同特征参数的相对偏差幅值和偏差极性分布规律判别各缸工作状态并定位状态异常缸。

Recognition of Cylinder Working Condition in Internal Combustion Engine Based on Multi-feature Fusion of Instantaneous Speed

In order to identify the working condition of each cylinder of the internal combustion engine by analyzing the corresponding relation of the instantaneous speed fluctuation process and each cylinder work phase,an accurate test method of the internal combustion engine instantaneous speed based on the hardware counting principle is proposed.The tests are carried out on a self-propelled gun engine,and the instantaneous speed signals were measured in the conditions of normal work,incomplete combustion and misfire.Fourier analysis was done to the signal,the electronically-controlled component and the fluctuation component caused by each cylinder acting impulse are separated by filtering according to different frequency bands,and then five characteristic parameters are extracted.Then,the relative deviation of each characteristic parameter and its polarity are calculated.Finally,the working condition of each cylinder and the position of the abnormal cylinder are distinguished according to the amplitude of each relative deviation and its polarity distribution regular rule.