立井刚性罐道典型故障的模式识别研究

发布时间：2018-03-23 02:05

本文选题：罐道典型故障　切入点：模式识别　出处：《中国矿业大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着煤矿生产年限的增加,刚性罐道会出现接头错位、局部凸起、整体倾斜等典型故障。罐道的状态对矿井提升系统工作的稳定性和安全性有着十分重要的影响。因此,优化立井刚性罐道的测试方法,对采集得到的信号提取故障特征参数,选择有效的方法对罐道的典型故障进行模式识别研究对提升机安全运行具有重要意义。本论文建立了提升钢丝绳-提升容器及刚性罐道-提升容器的动力学模型并进行仿真分析,在实验室环境下模拟了刚性罐道的典型故障,采用小波变换模极大值及经验模态分解的方法提取了罐道故障的特征向量,最后利用支持向量机法完成了立井刚性罐道典型故障的模式识别研究。具体内容包括:1)建立了提升钢丝绳-提升容器的垂直振动模型,分析了钢丝绳动位移对提升容器垂直方向加速度的影响;建立了提升容器水平方向多自由度动力学模型,利用Matlab的Simulink功能对典型故障下的输出响应进行了仿真,对提升容器振动加速度幅值的影响因素进行了分析;2)构建了基于虚拟仪器技术的刚性罐道典型故障信号的测试系统,分别设置了正常状态、接头错位、局部凸起及罐道倾斜故障,在不同实验对比条件下采集了提升容器的振动响应信号和倾角输出信号;3)采用自适应冗余第二代小波软阈值函数对振动信号进行降噪处理,针对采集信号的具体特征,对提升容器的振动响应信号采用小波变换模极大值和包络分析法提取故障特征,对提升容器的倾角输出信号采用经验模态分解法取其趋势项作为故障特征;4)以刚性罐道典型故障的四种模式:正常状态、接头错位、局部凸起及整体倾斜为对象,采用支持向量机的方法进行模式识别研究,获得各状态下模式识别的识别率。对识别的结果进行分析总结,说明该方法能够有效地实现立井刚性罐道典型故障的模式识别。
[Abstract]:With the increase of coal mine production life, typical faults such as joint dislocation, local protruding and integral tilt will appear in rigid tank path. The state of tank path has a very important effect on the stability and safety of mine hoisting system. The test method of shaft rigid tank channel is optimized, and fault characteristic parameters are extracted from the collected signals. It is of great significance to select effective methods for pattern recognition of typical faults of tank path for safe operation of hoist. In this paper, the dynamic models of hoisting wire rope hoisting vessel and rigid canal-lift vessel are established. And carry on the simulation analysis, The typical faults of the rigid tank track are simulated in the laboratory environment. The eigenvector of the fault is extracted by using the wavelet transform modulus maximum and empirical mode decomposition method. Finally, the support vector machine (SVM) method is used to complete the research on the pattern recognition of the typical faults of the shaft rigid tank path. The specific contents include: 1) the vertical vibration model of the hoisting wire rope and lifting vessel is established. The influence of wire rope dynamic displacement on vertical acceleration of hoisting vessel is analyzed, and the dynamic model of horizontal multi-degree-of-freedom (MDOF) of hoisting vessel is established, and the output response of typical fault is simulated by using Simulink function of Matlab. The influence factors of vibration acceleration amplitude of hoisting vessel are analyzed. (2) A test system of typical fault signal of rigid tank track based on virtual instrument technology is constructed, and the normal state and dislocation of joint are set up respectively. The vibration response signal and dip output signal of the lifting vessel are collected under different experimental conditions. The adaptive redundant second generation wavelet soft threshold function is used to reduce the noise of the vibration signal. According to the specific characteristics of the collected signal, the wavelet transform modulus maximum and envelope analysis are used to extract the fault features of the vibration response signal of the lifting vessel. Empirical mode decomposition (EMD) method is adopted to analyze the output signal of inclined angle of lifting vessel. The four modes of typical faults of rigid tank track are taken as its trend term. The object is normal state, dislocation of joint, local protruding and overall tilt. Support vector machine (SVM) is used to study pattern recognition, and the recognition rate of pattern recognition is obtained. The results of recognition are analyzed and summarized, which shows that this method can effectively realize the pattern recognition of typical faults of vertical shaft rigid tank track.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TD531

【参考文献】