摘要
文章设计一种基于自学习理论的机械故障检测方法,该方法通过对多组机械振动信号的采集,由SVM训练器进行故障的初步检测,经过多样本投票算法实现故障检测的最终结果判定。该监测方法综合应用了SVM分类器,以及多样本采样检测并投票的思路,提高了机械故障检测的精度。
Article design a mechanical fault detection method based on self-learning theory, the method of multiple sets of mechanical vibration signal acquisition, initial fault detection performed by the SVM trainers, through various voting algorithm to achieve the end result of this failure detection determination. The integrated application monitoring method SVM classifier, and detecting and voting varied ideas this sampling, improve the accuracy of a mechanical fault detection.
出处
《大众科技》
2015年第7期66-68,共3页
Popular Science & Technology
基金
广西教育厅科研课题"基于自学习理论的机械故障诊断技术研究"(KY2015LX652)
关键词
自学习理论
故障诊断
SVM
Since learning theory
fault diagnosis
SVM
