论文部分内容阅读
图像处理技术在内燃机的故障诊断与状态监测中应用还很少 ,本文探讨了应用于这方面时应注意的几个问题。指出利用缸盖振动信号小波包分解后的时 -频分布图的灰度直方图进行故障诊断的效果并不好 ,并分析了原因 ;进行了信号加噪声与不加噪声的诊断效果对比 ,发现噪声对基于图像处理的气门机构故障诊断的影响不大 ,验证了这一方法的工程实用性和可用性。
Image processing technology is rarely used in the fault diagnosis and condition monitoring of internal combustion engine. This article discusses several issues that should be paid attention to in this aspect. It is pointed out that the fault diagnosis using the histogram of the time-frequency distribution of the vibration signal of the cylinder head after the wavelet packet decomposition is not effective and the reason is analyzed. The comparison of the diagnostic results between the signal plus no noise and the noise is made. Noise has little effect on fault diagnosis of valve train based on image processing, which verifies the engineering practicability and usability of this method.