峰度
情态动词
模式识别(心理学)
计算机科学
信号(编程语言)
断层(地质)
方位(导航)
支持向量机
混淆矩阵
随机森林
分类器(UML)
算法
数学
人工智能
统计
地质学
高分子化学
程序设计语言
化学
地震学
作者
Heyu Zhang,Yuqiao Zheng,Jieshan Lu
摘要
A new fault diagnosis approach based on bearing current signals is proposed in this paper. First, in view of strong background noise of the current signal, the variational modal decomposition method is applied to decompose the bearing current signal to obtain multiple intrinsic mode functions, and then the intrinsic mode functions are constructed as the input feature vector according to the kurtosis. Second, to avoid the influence of random forest parameters on the random forest classifier, a random forest faulty bearing diagnostic model optimized by the whale algorithm is established. Finally, the accuracy rate and confusion matrix are adopted to evaluate the prediction effects of both established and traditional models. The classification accuracy of the real damaged bearing fault type can reach 95.11%. The fault diagnosis accuracy of manually damaged bearings can reach 93.83%. The results show that the method proposed in this paper has high accuracy and good generalization ability for bearing fault diagnosis.
科研通智能强力驱动
Strongly Powered by AbleSci AI