随着机械装备服役时间的增加,各关键部件渐渐进入耗损失效期,部分关键部件会发生轻微故障,如果不能及时准确识别这些微弱故障,会导致微弱故障劣化为严重故障,造成经济损失和机毁人亡的严重后果.如果能在这些故障程度尚轻微的阶段就能准确地识别出来,并及时掌握关键功能部件的健康状况,对防止轻微故障的劣化发展具有重要意义.以机车齿轮箱为研究对象,针对多信号受到不确定性因素影响而存在微弱性和混叠性难以识别的问题,提出一种有效抑制模态混叠现象并且实现有效获取故障特征频率的动态级联经验模态分解方法,通过改进包络线的求解方法获取最优本征模态函数,使故障特征在不同时间尺度上表现更加明显,实现在大量噪声背景下的微弱故障特征获取,为及时、准确进行早期故障预示及诊断提供一种有效手段.%Many key parts of mechanical equipment enter failure period gradually with the increase of service time,with minor faults occuring in some key components.Failure to identify the minor faults timely and accurately will lead to serious faults,which may cause serious consequences of economic losses,equipment damage and casualties.Accurate identification of the faults during the minor fault stage,and timely monitoring of the conditions of key parts will be of great significance to prevent the deterioration of the minor faults.With locomotive gearbox as the research object,in response to the difficulty in identifying the multi-signal problems in tenuity and aliasing under uncertain factors,a method was presented to effectively inhibit mode mixing and acquire fault characteristic frequency based on dynamic cascade empirical mode decomposition.The optimal intrinsic mode function was obtained by improving the envelope solution to exhibit fault features more obviously at different time scales.The results of the experiment showed that weak fault features can be obtained under the condition of much noise,which provides an effective method for early fault prognosis and diagnosis.
展开▼
