基于CNN的炉子集中润滑站润滑脂泄漏检测

Lubricating Grease Leakage Detection of Furnace Centralized Lubrication Station Based on CNN

退火炉是冷轧带钢生产线的主要工艺设备,目前,国内多采用双线干油润滑系统(集中润滑站)保证炉辊轴承的润滑效果。为了提升集中润滑站的工作可靠性,采用CNN检测方法进行润滑管路的润滑脂泄漏检测。通过收集管路多个位置漏油的图片,划分样本,构建基于CNN的润滑脂泄漏检测的图像识别二分类模型,训练并保存最优模型,驱动最优模型实时检测集中润滑站。该方法可提升集中润滑站管路的点检效率,提升可靠性,为退火炉设备的智能健康管理提供新思路。

Aannealing furnace is the main process equipment of cold rolled strip production line.Currently,a dual line dry oil lubrication system(centralized lubrication station)is commonly used in China to ensure the lubrication effect of the furnace roller bearings.In order to improve the reliability of the centralized lubrication station,this article uses CNN to detect grease leakage in the lubrication pipeline.Firstly,collect images of oil leaks from multiple locations in the pipeline and divide the samples.Then construct an image recognition binary classification model for lubricating grease leakage detection based on CNN.Finally,train and save the best model to drive the optimal model to detect the centralized lubrication station in real-time.This method can improve the efficiency of centralized lubrication station pipeline inspection,enhance reliability,and provide new ideas for intelligent health management of annealing furnace equipment.