基于奇异熵的磨粒群特征提取

为定量分析磨粒群,利用销-盘摩擦磨损试验机进行船用柴油机活塞环-缸套材料配副磨合磨损试验.在铁谱显微镜上对不同磨损阶段的磨粒群图像进行观察分析,应用奇异熵提取磨粒群图像的低频特征参数和高频特征参数.结果表明:低频特征参数和高频特征参数刻画了磨粒群图像上磨粒的形态特征.低频特征参数反映了磨粒群图像上磨粒尺度大小,其值越大,磨粒越大;高频特征参数反映了磨粒群图像上磨粒的数量,其值越大,磨粒越多.低频特征参数和高频特征参数可作为磨粒群特征参数.

Feature extraction of wear particle group based on singular entropy

Running-in wear test for materials of piston ring against cylinder liner was conducted by a pin-on-disc tester.The wear particle group image under different rubbing status was analyzed by a ferrographic microscope.Low frequency feature parameters and high frequency feature parameters of wear particle group image were extracted by using singular entropy.The results show that both low frequency feature parameters and high frequency feature parameters reveal the feature of wear particle.Low frequency feature parameters reflect scale of wear particle,the larger low frequency feature parameters,the larger wear particle.High frequency feature parameters reflect quantity of wear particle,the larger high frequency feature parameters,the more wear particle.Low frequency feature parameters and high frequency feature parameters can be considered as feature parameter of wear particle group.