摘要
为了解决机电设备早期故障难以正确识别及故障发展状态不易准确监测的问题,提出了一种基于模糊支持 矢量数据描述(FSVDD)的早期故障智能监测诊断新方法。该方法只需要一类目标样本作为学习样本就可以建立起 单值分类器,同时在核函数中引入非目标样本的模糊隶属度,从而把非目标样本与目标样本分等级地区分开来。 将这种方法应用在机电设备状态监测和故障诊断中,只需要将正常运行时的数据信号作为目标样本,就可以实现 对设备早期故障的准确识别,同时判断故障的严重程度。在轴承运行状态监测中的测试结果表明,该方法不仅能 快速识别轴承的早期故障,而且可以对故障的严重程度做出准确的判断。
In order to solve the problems of correctly identifying incipient fault and accurately monitoring fault development for electromechanical equipment, a new method of incipient fault intelligent monitoring and diagnosis based on fuzzy support vector data description(FSVDD) is proposed. With this method, one-class classifier can be built when only the information of the target class is available, and the outlier objects can be hierarchically distinguished from target objects when these membership degrees of outlier objects are appended to the kernel function. The proposed method is applied to the condition monitoring and fault diagnosis of electromechanical equipment, which can detect incipient fault only using normal condition signals and identify the fault severity. The experimental result shows that this method not only fast detects the bearing incipient fault, but accurately identifies the fault severity.
出处
《机械工程学报》
EI
CAS
CSCD
北大核心
2005年第12期145-150,共6页
Journal of Mechanical Engineering
基金
国家自然科学基金重点(50335030)国家自然科学基金(50175087
50305012)国家重点基础研究发展计划(973计划)(2005CB724106)高校博士点基金(20040698026)资助项目。
关键词
模糊支持矢量数据描述
单值分类
早期故障
智能监测诊断
Fuzzy support vector data description One-class classification Incipient fault Intelligent monitoring and diagnosis