摘要
针对传统滤波器对齿轮箱信号去噪不充分和模型识别准确率低的问题,提出一种基于变分模态分解(variational mode decomposition,VMD)和改进麻雀搜索算法(sparrow search algorithm,SSA)来优化极限学习机(extreme learning machine,ELM)的齿轮箱故障诊断模型。通过改进VMD后含噪分量的选取方式,并结合小波包阈值处理对齿轮箱信号进行滤噪,在提取时频域有效特征的基础上,通过Tent混沌映射和引入微分递减因子改进SSA以优化ELM模型进行分类识别。实验结果表明,所提模型对齿轮箱故障工况的分类准确率达到99.50%,在故障诊断精度提升的同时收敛速度更快,验证了模型的可行性。
To address the problem of inadequate noise removal by traditional filters and low accuracy of model recognition,a gearbox fault diagnosis model of extreme learning machine(ELM)was proposed based on variational mode decomposition(VMD)and improved sparrow search algorithm(SSA).The gearbox signal was filtered out through improved selection of noisy components after VMD and wavelet packet threshold processing.Based on the extracted effective features in a time-frequency domain,SSA was improved by the Tent chaos mapping with the introduction of differential decrement factor,which optimized the ELM for classification recognition.The experimental results showed that the classification accuracy of the proposed model for gearbox fault achieved 99.50%,and the convergence speed was faster while the fault diagnosis accuracy was improved,which verified the feasibility of the model.
作者
孟博
郇战
时文雅
余中舟
周靖诺
王佳晖
MENG Bo;HUAN Zhan;SHI Wenya;YU Zhongzhou;ZHOU Jingnuo;WANG Jiahui(School of Microelectronics and Control Engineering,Changzhou University,Changzhou 213164,China;Aliyun School of Big Data,Changzhou University,Changzhou 213164,China;Jiangsu Lida Elevator Co.,Ltd,Changzhou 213300,China)
出处
《郑州大学学报(理学版)》
CAS
北大核心
2024年第2期80-86,共7页
Journal of Zhengzhou University:Natural Science Edition
基金
国家自然科学基金项目(61772248)。
关键词
齿轮箱故障诊断
变分模态分解
小波包去噪
Tent混沌
麻雀搜索算法
极限学习机
gearbox fault diagnosis
variational mode decomposition
wavelet packet denoising
Tent chaos
sparrow search algorithm
extreme learning machine