【作者】 勾轶；

【导师】 陈长征；

【作者基本信息】 沈阳工业大学 ， 固体力学， 2004， 硕士

【摘要】 随着经济建设的发展和电气化程度的提高，电机已被广泛应用于工业生产的各个领域，电机一旦发生故障，即使是停止工作的时间极短，也会造成局部或大区域的停电，从而造成巨大的经济损失和重大的社会影响。因此，研究不同场合、不同运行状态下电机故障诊断理论和技术是提高生产设备可靠运行的保证。电机故障的类型多种多样：既有缓变故障和突变故障，又有电气故障和机械故障；既有线性系统故障，又有非线性系统故障，其关系错综复杂。诊断技术就是要通过各种检测技术，测定出能反映故障隐患和趋向的参数，从中得到预警信息。进一步通过信息分析对电动机故障程度和起因有一个准确判断，能及时和有效的对电动机进行维修、排除故障，以实现电动机的预知维修，而不致影响生产。 本文基于小波分析对中小型电机故障诊断进行研究，同时应用Fourier分析，实现对电机电气故障和机械故障的综合诊断。 研究了基于小波分析的常用降噪方法并对各种降噪方法进行了比较，比较结果说明根据各频带携带能量进行降噪和浮动阈值法对电机信号降噪都能得到较好的降噪效果。而前一种方法需要大量的先验知识，采用浮动阈值法则不需要这些先验知识，同时能取得较好的降噪效果。 文中阐述了电机各种故障的机理、产生原因和其在电机振动频谱或电流频谱中的表现。利用小波包算法良好的时频分析能力，对750w分数槽化纤电机和22kw永磁电机进行故障诊断。首先应用浮动阈值法对原始信号进行降噪，实际测试表明取得了良好的效果；然后用FFT算法对电机故障进行初步诊断，在对难以用FFT变换判定的频段进行小波分析准确判断电机故障。实验测得22kw永磁电机的故障只有电气故障：气隙偏心；而750w分数槽化纤电机出现了转子断条和电机前端轴承外套损坏故障，为电机的电气故障和机械故障同时出现。结果表明应用小波算法可以准确识别多模式并存故障。更多还原

【Abstract】 With the development of economy construction and the improvement of electric degree, electric machinery has been used in many fields of industry. When electric machinery is in fault conduction fault, power cut will happen in some places so that there is huge economy loss and important influence on society. So the study on the fault diagnosis theory and technology of electric machinery under all kinds of instance is a pledge, which makes equipments work normally. There are many kinds of electric machinery. For example, slow variety fault and sudden variety fault, electric fault and machinery fault, linearity system fault and unlinearity system fault, their ties are very complex. Fault diagnosis can get the information of beforehand alarm through the parameters, which can reflect the hidden trouble and trend measured by all kinds of technology. Furthermore, we draw a correct conclusion of the diagnosis degree and origin, so we can maintain electromotor in time and get rid of the diagnosis. And we can realize predict service and don’t influence the produce.Through the Wavelet analysis and the study of wavelet analysis we can realize comprehensive diagnosis of electric fault and machinery fault.In the paper the mechanism, cause and exhibition in vibration frequency and current frequency of electric machinery are stated. Fault diagnosis is made in 750w fraction slot melt electric machinery and 22kw permanent magnet electric machinery with good frequency ability of wavelet analysis. Firstly, db4 wavelet reduction noise of original signal ,which shows the testing is good; secondly, FFT analysis does with the fault of electric machinery accidentally, and those which can’t be deal with FFT are dealt with wavelet analysis on electric machinery. The result from the experiment of 22kw permanent magnet electric machinery shows that it only has electric fault: gas interval eccentricity; 750w fraction slot melt electric machinery has break of rotor twig and beaming bear coat shatter diagnosis, which is electric fault andmachinery fault. The examination shows that wavelet analysis can identify many faults at the same time veraciously.更多还原