模糊函数图像在柴油机气阀故障诊断中的应用研究

Study on the Use of Ambiguity Function Images in Fault Diagnosis for Diesel Valve Train

计算了柴油机气阀机构 8种状态下的缸盖表面振动信号的模糊函数 ,将结果在频偏—时延相平面上用灰度图表示出来 ,得到了一系列模糊函数图像。选取一部分模糊函数图像进行平均 ,得到了各种状态的标准模糊函数图像 ,再根据图像之间的欧氏距离、相似度和J 散度等 6种指标对模糊函数图像进行分类 ,从而将气阀机构的故障诊断转换为模糊函数图像的分类识别。试验结果表明 ,利用模糊函数图像可以取得很好的诊断结果 ,6种指标中欧氏距离和相似度两种指标的抗干扰能力比较强 ,更适合于作为模糊函数图像的分类指标。图像平均可以有效地提高故障诊断的正确率。

The ambiguity functions of eight kinds of vibration acceleration signals, which were acquired from the cylinder head in eight different states of valve train, were calculated and displayed in grey-level images in this paper. A series of ambiguity function images were obtained, from which eight standard ambiguity function images were built. By using six indexes, such as euclidean distance, similarity, J-divergence and so on, the testing images were classified into eight types, which correspond to eight states of valve train. The process of fault diagnosis for valve train was transformed to the classification of ambiguity function images. The experimental results showed that a very high rate of correct recognition could be obtained using ambiguity function images. Among the six indexes, Euclidean distance and similarity are more suitable for classifying. The method of Image averaging can improve the correct recognition rate of fault diagnosis effectively.