|
| 基于随机森林的电机异音故障诊断方法 |
| Fault Diagnosis Method of Motor Abnormal Sound Based on Random Forest |
| |
| DOI：10.16018/j.cnki.cn32-1650/n.202302007 |
| 中文关键词: 电机 振动信号 异音识别 随机森林 主成分分析 特征提取 |
| 英文关键词: motor vibration signal abnormal sound recognition random forest principal component analysis feature extraction |
| 基金项目: |
|
| 摘要点击次数: 44 |
| 全文下载次数: 62 |
| 中文摘要: |
| 针对电机异音故障检测技术存在准确率低、模型复杂等问题，提出一种基于随机森林的电机异音故障诊断方法。通过自行研制的汽车智能座椅靠背电机振动测试平台，分析电机故障产生过程及异音的特征，并从时域中提取11个特征表征异音信号的变化；通过主成分分析法对所提取的特征进行降维，在训练基于随机森林和概率神经网络的电机故障智能识别方法基础上，通过自行研制的汽车智能座椅靠背电机振动测试平台采集数据，得到随机森林的平均识别准确率为95. 11%±2. 17%，概率神经网络的平均识别准确率为93. 90%±2. 16%。 |
| 英文摘要: |
| Aiming at the problems of low accuracy and complex model of motor abnormal sound fault detection technology, a motor abnormal sound fault diagnosis method based on random forest is proposed. Based on the self-developed vibration test platform of automotive intelligent seat back motor, the process of motor failure and the characteristics of abnormal sound are analyzed, and 11 features are extracted from the time domain to characterize the changes of abnormal sound signals. The dimensionality of the extracted features is reduced by the principal component analysis method. On the basis of training the motor fault intelligent identification method based on random forest and probabilistic neural network, data is collected through the self-developed vibration test platform of automotive intelligent seat back motor, and the average recognition accuracy of random forest is 95. 11%±2. 17%. The average recognition accuracy of probabilistic neural network is 93. 90%±2. 16%. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |