基于多段特征提取的永磁直流电机故障诊断
Permanent Magnet DC Motor Fault Diagnosis Based on Multi-Segment Feature Extraction
李梦宇1
作者信息
- 1. 宝钛集团铸件材料公司有限公司,陕西宝鸡 721000
- 折叠
摘要
对于永磁直流电机(PMDCM),电流信号的振幅在电机启动后会逐渐减小,仅使用单段电流信号特征不利于永磁直流电机的故障诊断,为此,提出了多段特征提取方法,以提高永磁直驱单片机的故障诊断效果.此外,还利用支持向量机(SVM)、分类与回归树(CART)以及k近邻算法(k-NN)构建故障诊断模型.从多个连续的电流信号分段中提取的时域特征构成一个特征向量,用于PMDCM的故障诊断.结果表明,与单段特征相比,多段特征具有更好的诊断效果,故障诊断的平均准确率提高了19.88%;通过多区段特征提取为PMDCM的故障诊断奠定了基础,并提供了一种新颖的特征提取方法.
Abstract
For permanent magnet DC motors(PMDCM),the amplitude of the current signal gradually decreases after the motor starts.Using only single segment current signal features is not conducive to fault diagnosis of permanent magnet DC motors.Therefore,a multi-segment feature extraction method is proposed to improve the fault diagnosis effect of permanent magnet direct drive microcontrollers.In addition,support vector machine(SVM),classification and regression tree(CART),and k-nearest neighbor algorithm(k-NN)are used to construct a fault diagnosis model.The time-domain features extracted from multiple continuous current signal segments form a feature vector for fault diagnosis in PMDCM.The results show that compared with single segment features,multi-segment features have better diagnostic performance,and the average accuracy of fault diagnosis has increased by 19.88%.The multi-segment feature extraction lays the foundation for the fault diagnosis of PMDCM and provides a novel feature extraction method.
关键词
多段特征/提取/永磁直流电机/故障检测Key words
multi-segment features/extraction/permanent magnet DC motor/fault detection引用本文复制引用
出版年
2024
NETL
NSTL
万方数据