为了准确实现内燃机气门间隙异常的诊断,提出了基于频带能量特征向量时频域特征的气门间隙故障诊断方法。介绍了小波包改进算法原理及频段能量计算,并以S195四冲程柴油机为研究对象,测取气门间隙正常、进气门间隙异常和排气门间隙异常等不同工况下的缸盖振动加速度信号,并对其进行改进小波包分解处理,通过6~8 kHz高频带能量的故障阈值和分析信号时域内气门开启、关闭时刻对应的曲轴转角的变化确定进气门或排气门间隙异常,试验结果表明采用该方法诊断出各工况的故障与实际故障吻合。因此,通过小波包分解后各频带相对能量分布来诊断内燃机气门间隙故障是一种有效且可行的方法。
In order to gain an accurate diagnosis of abnormal engine valve clearance, the paper presents a valve clearance fault diagnosis method based on time-frequency domain characteristics of frequency band energy eigenvector. The principle of improved wavelet packet algorithm and the energy band calculation are introduced. Taking the S195 four-stroke diesel engines as an example, the cylinder cover vibration acceleration signal under the conditions of normal valve clearance, abnormal intake valve clearance and abnormal exhaust valve clearance are measured and processed by improved wavelet packet decomposition. The clearance of intake valve and exhaust valve can be determined through the 6 - 8 kHz high frequency band energy value of faulty valve and the analysis of the variation of crank angle corresponded by the open and close time of intake and exhaust valves within the signal time domain. The test results show that the diagnosis made by using the new method match well with the actual fault condition. Therefore, it is feasible and effective to make engine valve clearance fault diagnosis by the relative energy distribution of each frequency band decomposed by wavelet packet.