基于数据驱动的船舶旋转机械故障诊断方法研究及应用

发布时间：2018-03-20 03:06

本文选题：故障诊断　切入点：船舶旋转机械　出处：《南京航空航天大学》2016年硕士论文　论文类型：学位论文

【摘要】：旋转机械是工业领域应用最为广泛的一类机械设备,也是船舶系统的重要组成部分。由于受到工艺缺陷、安装不当以及长期恶劣的工作环境影响,船舶的旋转机械设备极易发生故障。这些故障不仅会影响船舶正常航运,还会引发重大经济损失和人员伤亡。当前船舶系统故障诊断多依赖简单的检测设备和专家经验,误诊率较高。因此,针对船舶的旋转机械设备,研究其故障诊断方法具有重要的理论价值和实际意义。本文结合工程背景,利用基于数据驱动的方法对船舶旋转机械进行故障诊断,工作主要包括理论分析和应用研究。本文开始概述了故障诊断方法,介绍了基于数据驱动的旋转机械故障诊断方法的发展历程,并分析了现有成果的特点与不足。首先,通过实船监测和实验方式获得了大量的研究数据。实船数据采集于航运中的散货轮,数据分为正常状态和故障状态两类。实验数据来源于船舶旋转机械故障模拟实验平台,包含了几种常见故障的数据以及正常状态下的数据。其次,分析了经验模式分解方法在信号分析中的应用,包括对传统信号降噪方法进行优化,以及结合样本熵理论实现船舶旋转机械的故障诊断。首先根据分量相关性原则,提出将遗传算法与经验模式分解阈值降噪方法相结合,增强降噪后信号与原信号的相关性。然后将经验模式分解与样本熵理论相结合,完成了船舶旋转机械的故障诊断。接着,针对传统信号分析方法去噪难的缺陷,提出了基于灰度图纹理分析的船舶旋转机械故障诊断方法。将船舶旋转机械振动信号转换为灰度图像,利用相应的纹理图像分析方法完成故障识别,并从理论和实验两个方面验证了该方法的可行性和实用性。最后,进行了基于Matlab图形用户界面的船舶旋转机械设备故障诊断系统开发。介绍了系统的开发环境、总体结构设计,说明了各模块的实现方法,并测试了该软件系统的有效性和实用性。
[Abstract]:Rotating machinery is one of the most widely used machinery and equipment in industrial field. It is also an important part of ship system. The ship's rotating machinery is prone to malfunction. These faults will not only affect the normal shipping of the ship, but also lead to heavy economic losses and casualties. At present, the fault diagnosis of ship system depends on simple detection equipment and expert experience. Therefore, it is of great theoretical value and practical significance to study the method of fault diagnosis for rotating machinery and equipment of ships. The work of fault diagnosis of ship rotating machinery based on data-driven method mainly includes theoretical analysis and application research. In this paper, the method of fault diagnosis is summarized. This paper introduces the development course of fault diagnosis method for rotating machinery based on data drive, and analyzes the characteristics and shortcomings of existing achievements. A large number of research data are obtained by means of real ship monitoring and experiment. The data are collected from bulk cargo ships in shipping, and the data can be divided into two types: normal state and fault state. The experimental data are derived from the fault simulation experiment platform of ship rotating machinery. It contains several kinds of common fault data and data under normal condition. Secondly, the application of empirical mode decomposition method in signal analysis is analyzed, including the optimization of traditional signal denoising methods. According to the principle of component correlation, the genetic algorithm is combined with empirical mode decomposition threshold to reduce noise. The correlation between the noise reduction signal and the original signal is enhanced. Then the empirical mode decomposition is combined with the sample entropy theory to complete the fault diagnosis of ship rotating machinery. A fault diagnosis method for ship rotating machinery based on grayscale image texture analysis is proposed. The vibration signal of ship rotating machinery is converted into gray image, and the fault identification is accomplished by using the corresponding texture image analysis method. The feasibility and practicability of the method are verified theoretically and experimentally. Finally, the fault diagnosis system of marine rotating machinery equipment based on Matlab graphical user interface is developed. The development environment and overall structure design of the system are introduced. The realization method of each module is explained, and the validity and practicability of the software system are tested.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：U672

【相似文献】