在线铁谱仪已成功地用于机械设备的磨损状态监测,如何根据铁谱数据来分析机器磨损状态和磨损规律一直被人们所重视.本文利用自组织特征映射神经网络对齿轮整个磨损过程的铁谱数据进行分类,提出了齿轮不同磨损状态下磨粒的分布规律,根据这种规律可以识别齿轮的不同磨损状态.这对监测机械设备运行状态具有十分重要的意义. On-line iron spectrometers have been successfully used to monitor wear status of machinery and equipment, and how to analyze wear status and wear rules of machines based on iron spectrum data has been paid attention to by people.In this paper, self-organizing feature mapping neural network According to the law, the different wear states of gears can be identified, which is very important for monitoring the running status of mechanical equipment.