【作者】 石磊；

【导师】 向阳；

【作者基本信息】 武汉理工大学 ， 轮机工程， 2012， 硕士

【摘要】 船舶主机是船舶的动力心脏,其安全稳定的运行是保证船舶航行安全的关键。海洋救助船的作用是救助在航行中出现故障的船舶,特别是出现动力失效故障的船舶,遇到呼救能够及时执行任务是对救助船的基本要求,所以救助船对于机舱动力设备的可靠性要求很高。本论文的目的就是研究一套合理而有效的海洋救助船主动力装置监测诊断方法,并设计一款主动力装置监测与故障诊断系统。完成的主要研究工作如下：(1)对齿轮箱故障诊断的振动方法进行了研究,对齿轮箱发生齿轮破损、断齿、轴不平衡等故障的振动信号进行时域分析,频域分析以及时频分析等,提取关键特征值,分析不同故障状态时特征值的大小与变化,重点研究了不同故障工况振动信号时域图与频域图中特征的变化。(2)以6L16/24柴油机为研究对象,对其瞬时转速的特性进行了研究。应用多种方法提高了瞬时转速的提取精度,准确还原了瞬时转速波动信号,分析了正常与故障状态的瞬时转速信号,并由此提取瞬时转速特征值,基于这些特征值确定实时在线监测柴油机故障的瞬时转速诊断方法。(3)选取了部分齿轮箱振动特征值作为神经网络特征向量输入,利用神经网络进行了齿轮箱故障模式识别,从训练时间,稳定性,识别成功率等方面比较了RBF网络与BP网络进行齿轮箱模式识别的效果,研究了不同参数对网络训练性能的影响。结果表明,该方法能够准确识别齿轮箱故障状态,RBF训练速度快,识别效果好。(4)以Lab VIEW虚拟仪器为工具,以上述理论研究为基础,设计了8000kW海洋救助船主动力装置监测诊断系统。系统由热力参数监测诊断模块、瞬时转速监测诊断模块、示功图监测诊断模块、齿轮箱监测诊断模块以及融合诊断模块组成。更多还原

【Abstract】 Diesel engine is the core part of vessel, the safety of main engine ensures the safety of the vessel. Ocean salvage vessel’s role is help the ship which is break down, especially which loses power. For rescue ship quickly carry out the rescue task is significant, then the reliability of main power plant in salvage vessel should be high. In the thesis, the on-line monitoring and fault diagnosis method of main engine and gear box is analyzed, then the system is designed. In details, the main work is as follows:(1)Research on the gear box vibration mechanism. The vibration signal of typical failures including chipped tooth, missing tooth, shaft imbalance are analyzed using the method like time-domain, frequency-domain, Time-frequency analysis. The key values of the vibration signal in different conditions are extracted and analyzed. The time-domain waveform and frequency-domain waveform are mainly studied.(2)Research on the instantaneous angular velocity waveform of6L16/24main engine, the velocity waveform is extracted accurately by some theory and signal processing method. The instantaneous angular velocity signal in normal and fault condition are analyzed, key value of the velocity waveform is extracted. On the basis of these key values, the instantaneous angular velocity waveform method in main engine fault diagnosis is summarized(3)The principle and classification of Artificial Neural Network are studied, the principle,structure and train method of the RBF and BP network are mainly analyzed, some key values are chose as neural network inputs for pattern recognition of different failures in gearbox. The result of the recognition between RBF and BP is compared; the parameters which influence network performance are studied. The results indicated that ANN can accurately recognize the condition of gearbox, compared with BP network, RBF network is better in train time and recognition result.(4)Based on the work mentioned above, the on-line monitoring and fault diagnosis system of main power plant in8000kW ocean salvage vessel is designed. The system based on LabVIEW and consists several parts:Thermal parameter, instantaneous angular velocity, indicator diagram, gearbox vibration and fusion fault diagnosis modules.更多还原