【作者】 刘冬；

【导师】 蒋兴浩；

【作者基本信息】 上海交通大学 ， 电子与通信工程， 2010， 硕士

【摘要】 旋转机械作为一种重要的动力装置,广泛应用于发电,舰船,航空等领域,在国民经济中起着极其重要的作用。随着旋转机械机组越来越复杂,对其性能和状态的检测及控制也变得越来越重要。若发生故障,往往会产生错误的监测信号,带来不当的控制等非常严重的后果,因而旋转机械的故障诊断成为了一个比较重要的研究课题。旋转机械的故障诊断是一种了解和掌握机器在运行过程的状态,确定其整体或局部状态是否正常,早期发现故障及其原因,并能预报故障发展趋势的技术。本文主要针对基于振动信号处理的故障诊断进行了相关的研究工作:(1)利用实验室设备,进行了振动实验,采集了不同类型的实验数据,可为仿真计算提供数据。(2)对信号处理的基本流程进行了介绍,以振动信号为例,进行了信号处理的具体举例说明。(3)基于小波分析对振动信号进行了信号处理。基于小波分析在时频分析方面的优势,经过仿真计算,进行了小波基及小波分解尺度的选择,得出了小波变换在振动故障诊断处理方面的可行性。(4)基于美国国家仪器公司(National Instrument, NI)的硬件系统和软件系统,根据前面的理论知识和仿真计算,搭建了振动信号处理和故障诊断平台。这个平台为理论提供了有效的支撑和说明,同时也为后续的改进提供了有力的工具和条件。综上,本课题通过背景研究、理论分析、仿真计算、平台搭建以及实验验证的整套研究,提出了基于振动信号处理的旋转机械故障诊断的框架与方法,同时展望了该方法的发展趋势和方向,在工程应用上具有较大的意义和前景。更多还原

【Abstract】 Rotating machine, a prime mover that widely used in power generation, ship propulsion, aviation and other fields, plays a very significant role in the national economy. The status monitoring and the fault diagnosing system are becoming more and more important with the increasing complexity of rotating machine units. Once a failure occurs, the error signal produced would lead the rotating machine engine to a false status that could result in serious consequences. Therefore, the health state detection and fault diagnosis of rotating machine have become an important research topic.Fault diagnosis for rotating machine is a technology to understand the state of the machine during operation to determine the overall or partial normal or abnormal,to find the reasons for the early detection of failure, and to forecast the development trend of fault.In the paper, the following work of vibration signal is done:(1)Laboratory equipment is used to conduct a vibration testing, the different types of experimental data are collected which provides data for simulation.(2)The basic process of signal processing were introduced, the vibration signal is taken for example to carry out signal processing in specific.(3)Wavelet analysis of vibration signals is carried out based on the the advantages of wavelet analysis in time-frequency analysis. Through simulation, wavelet base and scales of decomposition is chosen, and finally the feasibility of wavelet analysis for vibration signal processing is proved.(4)Based on National Instruments (National Instrument,NI) hardware and software systems,according to previous theory and simulation, vibration signal processing platform is build. This platform provides an effective support for theory,as well as to provide a powerful tool and condition for improvement.In summary,a fault detection and diagnosis framework and method is proposed for vibration signal through background study,theoretical analysis, simulation and calculation,platform building and experimental verification. The development trend and prospect of the method is also studied and it is concluded that the method is of great significance and prospects in engineering applications.更多还原