磨粒直接成像技术在设备润滑状态监测领域的应用

Application of Direct Wear Debris Laser Imaging Techniques in Lubrication State Monitoring for Equipment

润滑油中的磨粒是判断设备是否正常运转以及失效原因的重要依据。研究一种基于人工神经网络和直接成像技术进行磨粒分析的新方法,该方法结合颗粒计数和磨粒分类两方面的信息,不仅能通过磨粒的数量、尺寸、尺寸分布等,检测设备是否正常运转及磨损程度,而且可以根据磨粒磨损类型诊断设备故障的可能原因。相比与其他方法,该方法检测速度快,可广泛应用于离线实验室和现场在线油液检测。该方法已成为最新的ASTM D7596标准。

The wear debris present in a lubricating oil is an imporatnt basis for determining whether that equipment is operating properly and the causes of failure. A new method is developed to use direct laser imaging techniques and ad- vanced neural network specifically to identify the wear debris.This method was combined the information of both particle counting and wear shape classification, it can determine the type, rate of production, and severity of mechanical faults by measuring the size, the size distribution and quantity of wear debris, and also can diagnose the possible root cause of the problem from wear particle and shape classifications.As compared with other methods this method is fast in testing,which can be widely applied for offline laboratory and oil testing online.The method has already become the newest standard of ASTM D7596.