基于证据理论多源多特征融合的柱塞泵故障诊断方法.pdfVIP

学兔兔 第9卷第1期 中 国 工 程 机 械 学 报 VO1．9 No．1 2011年3月 CHINESE JOURNAL OF CONSTRUCTION MACHINERY Mar．2011 基于证据理论多源多特征融合的 柱塞泵故障诊断方法 吴胜强L ，姜万录 (1．燕山大学 机械工程学院，河北 秦皇岛 066004；2．邢台职业技术学院，河北 邢台 054035) 摘要：提出了1种基于证据理论多源多特征融合的故障诊断方法，构建了柱塞泵的故障特征向量，分析了单源 多特征融合故障诊断，利用证据理论融合多个单源诊断结果，进行了多源故障诊断．对液压泵进行了试验．试验 结果表明：多源数据较单源数据，更具有冗余互补性，能提高诊断的可靠性和准确性．该诊断方法是可行而有 效的． 关键词：故障诊断；证据理论；时频域；多源多特征；柱塞泵 中图分类号：T 文献标识码：A 文章编号：1672—5581(2011)01—0098—05 Multi·source and multi-feature fused fault diagnosis on pi ston pumps based on evidence theory WUSheng-qiang 。JIANG Wan-lu (1．College of Mechanical Engineering，Yanshan University，Qinhuangdao 066004，China； 2．Xingtai Vocational and Technical Co llege，Xingtai 054035，China) Abstract：In this study，a multi—source and multi-feature integrated fault diagnosis method is first proposed based on evidence theory．In this regard，the fault feature vectors are extracted for piston pumps．Then，the evidence theory is applied for fusing the results from several single—source and multi- feature fault diagnosis．Finally，it is found from the testing results on hydraulic pumps that the multi— source data possess higher redundant complementarity than single-source data．Meanwhile，the feasibility and effectiveness of this approach is proven through enhancing the diagnosis reliability and accuracy． Key words：fault diagnosis；evidence theory；time—frequency domain；multi-source and multi—feature； piston pump 液压泵作为液压系统的动力源，是整个液压系统的心