基于磨粒磨损机理的机械磨损状态监测

针对机械设备磨损状态监测准确率较低的问题,基于不同磨损机理下磨粒具有不同的形状和纹理特征,提出了一种基于磨粒特征识别的机械磨损状态监测的数学模型。通过形状特征识别球状磨粒和切削磨粒,结合形状、纹理特征识别疲劳磨粒和严重滑动磨粒,基于提取的特征参数建立机械磨损状态监测的特征向量,通过量子粒子群优化(QPSO)的径向基函数神经网络模型,实现对机械磨损状态的监测和判别。实验结果表明:QPSO-RBF神经网络数学模型结构简单,比传统PSO-RBF神经网络模型的识别准确率高5%,可用于常见机械磨损状态的检测。 

Mechanical wear condition monitoring method based on abrasive particle wear mechanism

In order to solve the problem of low accuracy in monitoring the wear state of mechanical equipment, a mathematical model of monitoring the wear state based on the recognition of abrasive particle features was proposed based on different wear mechanisms with different shapes and textures. By identifying ball wear particles and cutting wear particles through shape feature, the shape and the texture features were combined to recognize fatigue wear particles and severe sliding wear particles. The feature vector of mechanical wear state monitoring was established based on the extracted feature parameters. Through the radical basis function neural network model by quantum particle swarm optimization (QPSO), the recognition and monitoring of mechanical wear state were realized. The experimental results show that the QPSO-RBF neural network model is simple in structure and 5% higher in recognition accuracy than the traditional PSO-RBF neural network model. It can be used for common mechanical wear condition monitoring