摘要
滚动轴承是电机的重要部件,及时、准确地对其进行故障诊断是电机安全运行的重要保障。针对滚动轴承常见的状态,包括正常、内圈轻微故障、滚动体轻微故障、外圈轻微故障、内圈中等故障、滚动体中等故障、外圈中等故障等七种情况,基于自组织神经网络,提出了电机轴承故障诊断的可视化方法。首先对采集到的振动信号进行特征提取,然后构建自组织神经网络,经过训练后,利用测试数据对诊断模型进行了测试,试验结果验证了所提方法的有效性。
Rolling element bearings constitute the key parts in motor, and the prompt and accurate fault detection of the bearings is very helpful in terms of enhancing the reliability of motor. For the common seven conditions, including normal condition, small inner race fault, small ball fault, small outer race fault, medium inner race fault, medium ball fanlt, and medium outer race fault, based on the self-organization map neural network, the visualization method of motor bearing fault diagnosis is presented. First, the features are extracted from the collected vibration signals. Then, the self- organization map neural network is constructed, and trained with the training samples. Last, the presented model is tested with the test samples. The experimental results show that the self-organization map neural network gets a promising result.
出处
《机械设计与制造》
北大核心
2013年第12期73-75,78,共4页
Machinery Design & Manufacture
基金
国家自然科学基金资助项目(51205371)
关键词
电机轴承
故障诊断
自组织神经网络
可视化
Motor Bearing
Fault Diagnosis
Self- Organization Map Neural Network
Visualization