基于优化SVM的轻载工况滚动轴承故障诊断  

Fault Diagnosis of Rolling Bearing under Light Load Operating Condition Based on Optimized Support Vector Machine(SVM)

作　　者：万庆 袁志鹏 王二振 曹朋 Wan Qing;Yuan Zhipeng;Wang Erzhen;Cao Peng(AVIC Hongdu,Nanchang,Jiangxi,330095)

机构地区：[1]中航工业洪都,江西南昌330095

出　　处：《教练机》2025年第1期63-68,共6页Trainer

摘　　要：特征提取是故障智能诊断的关键步骤,然而不同的特征提取方法所得到的特征不同,导致诊断结果也可能有所差异,且增加了人工特征选择的难度和不确定性。本文基于所搭建的试验台对不同径向载荷工况下正常状态、内圈故障、外圈故障轴承的振动信号进行了采集和分析,并针对轻载工况下滚动轴承故障诊断问题,结合支持向量机(SVM),利用遗传算法进行参数优化,通过原始数据实现轴承故障的分类识别。研究结果表明:不同径向载荷条件下故障信号表现出来的特征指标不同,在轴承故障诊断中使用传统的故障诊断会出现一定的误差;考虑径向载荷的影响、采用遗传算法优化的SVM故障诊断模型能够对故障类型进行更加有效的诊断,可提升准确率并降低计算成本。Feature extraction is a critical step in intelligent fault diagnosis.However,the features obtained by different feature extraction methods are different,which may lead to different diagnosis results,increasing difficulty and uncertainty of manual feature selection.This paper collects and analyzes the vibration signals of normal bearings and bearings with inner or outer ring failure under different radial load conditions based on the built test bench.In addition,to solve the problem of rolling bearing fault diagnosis under light load conditions,genetic algorithms are used for parameter optimization in conjunction with SVM,so as to realize the classification and identification of bearing faults through the raw data.The result of research shows that the characteristics of fault signal in different radial load conditions are different,and it will lead to certain errors by using traditional fault diagnosis.Considering the effects of radial load,SVM fault diagnosis model optimized by genetic algorithm is able to diagnose the fault types more effectively,which can also improve the accuracy and reduce the calculation costs.

关 键 词：轻载 滚动轴承 故障诊断 优化SVM 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] TH133.33[自动化与计算机技术—控制科学与工程]