基于灰度-梯度共生矩阵的往复压缩机气阀故障诊断方法研究

摘要

A gray level-gradient co-occurrence matrix-based extraction method for fault gas valve' s image was proposed, which taking smoothed pseudo Wigner-Vile distribution(SPWVD)of different gas valves as the object of study to obtain parameters of texture feature information extracted from the image. The results show that images reflect different parameters of the fault valve, this can have quantitative fault diagnosis of reciprocating compressor gas valve implemented.%提出基于灰度-梯度共生矩阵的气阀故障图像特征提取方法,以不同故障气阀阀片的平滑伪威格纳-维尔分布图为研究对象,通过提取图像中的纹理特征信息计算其纹理特征参数,由结果可知,不同的故障图像对应不同的特征参量,从而实现了往复压缩机气阀故障的定量诊断.