烧结主抽风机振动监测与故障诊断研究

发布时间：2019-01-05 20:09

【摘要】：主抽风机是烧结工艺中的关键设备，其运行状态直接关系着烧结生产的正常运行。对其实施状态监测和故障诊断可以保证主抽风机设备运行的安全可靠。传统基于傅里叶变换的信号处理方法是以平稳随机信号为前提，而小波分析作为现代信号处理技术，其良好时频局部化特性为处理的故障信号的非平稳问题提供了有效的数学工具。本文首先对烧结主抽风机设计了主抽风机设备振动的检测，为后面的故障诊断提供诊断依据。第二，提出了基于小波分析的模糊判决多故障分类器。主要研究了用小波分分析技术进行频带分析，即通过统计处理得到的典型故障状态的小波能量特征向量，建立故障诊断查询表。并针对故障诊断中的不确定信息，定义了故障类型模糊隶属度。模糊判决多故障分类器考虑了多故障并发时设备振动的差异，同时对主抽风机设备的轻微故障和早期故障诊断也具有较好的灵敏度。第三，改进了基于支持向量机的多故障分类器，并对支持向量机的参数进行了优化。针对SVM多类分类算法解决多类故障同时出现情况下故障识别的难题，将同时发生的两个故障看作新的故障类型，再通过支持向量机多分类算法进行故障识别。同时采用交叉验证、遗传算法和粒子群算法对支持向量机的参数寻优。最后，，设计了烧结主抽风机控制系统。应用RSLogix5000和RSView32软件，对烧结主抽风机设备网络数据通信、设备控制、故障诊断进行了分析，实现了烧结主抽风机状态实时监测和故障监测。
[Abstract]:The main extractor is the key equipment in sintering process, and its running state is directly related to the normal operation of sintering production. The condition monitoring and fault diagnosis can ensure the safety and reliability of the main ventilator equipment. The traditional signal processing method based on Fourier transform is based on stationary random signal, and wavelet analysis is a modern signal processing technology. Its good time-frequency localization property provides an effective mathematical tool for dealing with the non-stationary problem of fault signal. In this paper, first of all, the vibration detection of the main pumping machine is designed, which provides the diagnostic basis for the fault diagnosis. Secondly, a fuzzy decision multi-fault classifier based on wavelet analysis is proposed. In this paper, the wavelet energy eigenvector of typical fault state is obtained by using wavelet division analysis technique, and the fault diagnosis query table is established. According to the uncertain information in fault diagnosis, the fuzzy membership degree of fault type is defined. The fuzzy decision multi-fault classifier takes into account the difference of the vibration of the equipment when the multiple faults are concurrent and has good sensitivity for the minor faults and early fault diagnosis of the main ventilator equipment. Thirdly, the multi-fault classifier based on support vector machine is improved, and the parameters of support vector machine are optimized. In order to solve the problem of fault identification in the case of simultaneous occurrence of multiple types of faults, the SVM multi-class classification algorithm considers two simultaneous faults as a new fault type, and then uses support vector machine multi-classification algorithm to identify the faults. At the same time, the parameters of support vector machine are optimized by cross validation, genetic algorithm and particle swarm optimization. Finally, the control system of sinter main exhaust machine is designed. By using RSLogix5000 and RSView32 software, the network data communication, equipment control and fault diagnosis of sinter main exhaust equipment are analyzed, and the real-time monitoring and fault monitoring of sinter main air extractor are realized.
【学位授予单位】：辽宁科技大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TH165.3

【相似文献】