基于蚁群算法的离心式压缩机智能故障诊断方法研究
发布时间：2019-02-14 17:41
【摘要】：机械设备稳定高效的运行是保证生产安全和生产质量的重要前提,生产设备的动态监测和故障诊断具有极高的实际意义和研究价值,对设备准确的趋势预报、故障诊断和维修决策不仅可以降低设备的维护成本,更可以降低设备发生事故的风险。离心式压缩机作为一种重要的能量转化装置,在工业领域有着广泛的应用,它的工作状态和运行可靠性,对整个生产系统有着重要的影响。离心式压缩机作为一种旋转机械,其转子的工作状况对压缩机的稳定运行有较大的影响。造成离心压缩机故障的原因有很多种,加工误差、介质腐蚀、材料特性机械损伤、激振、颤振、喘振、磨损等都是影响离心压缩机正常工作的主要原因。本文中主要以油膜涡动、喘振、转子不平衡三种离心式压缩机典型故障为研究对象,通过对压缩机故障状态下径向振动信号进行分析,研究其振动特性,并以此作为故障诊断的依据,进一步研究蚁群算法在离心式压缩机故障诊断中的应用。蚁群式算法是一种启发式仿生算法,具有较强的模式识别能力,可对数据进行聚类分析。本文运用MATLAB建立蚁群算法模型,并进行模拟实验,对实测振动数据进行特征分析,通过对滚动轴承的故障诊断,验证蚁群算法在旋转机械故障诊断中的可行性,研究其诊断特性,并将诊断方法进一步优化延伸,对离心式压缩机实际故障工况进行分析和故障诊断,将诊断结果与实际情况进行对比,其诊断效果良好,程序运行稳定,结果准确。
[Abstract]:The stable and efficient operation of mechanical equipment is an important prerequisite to ensure production safety and production quality. The dynamic monitoring and fault diagnosis of production equipment is of great practical significance and research value. Fault diagnosis and maintenance decision can not only reduce the maintenance cost of equipment, but also reduce the risk of accidents. Centrifugal compressor, as an important energy conversion device, has been widely used in the industrial field. Its working state and operational reliability have an important impact on the whole production system. As a kind of rotating machinery, the working condition of centrifugal compressor rotor has great influence on the stable operation of compressor. There are many causes of centrifugal compressor failure, such as machining error, medium corrosion, mechanical damage of material characteristics, excitation, flutter, surge, wear and so on, which are the main reasons that affect the normal operation of centrifugal compressor. In this paper, three typical faults of centrifugal compressor, such as oil film vortex, surge and rotor unbalance, are studied. The vibration characteristics of centrifugal compressor are studied by analyzing the radial vibration signal of compressor under the condition of fault. As the basis of fault diagnosis, the application of ant colony algorithm in fault diagnosis of centrifugal compressor is further studied. Ant colony algorithm is a heuristic bionic algorithm which has strong pattern recognition ability and can be used for data clustering analysis. In this paper, MATLAB is used to establish ant colony algorithm model, and simulation experiments are carried out to analyze the characteristics of measured vibration data. Through fault diagnosis of rolling bearings, the feasibility of ant colony algorithm in fault diagnosis of rotating machinery is verified. The diagnosis characteristic is studied, and the diagnosis method is further optimized and extended, and the actual fault condition of centrifugal compressor is analyzed and diagnosed. The result of diagnosis is compared with the actual situation, and the diagnosis effect is good and the program runs stably. The result is accurate.
【学位授予单位】：西安石油大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TQ051.21
[Abstract]:The stable and efficient operation of mechanical equipment is an important prerequisite to ensure production safety and production quality. The dynamic monitoring and fault diagnosis of production equipment is of great practical significance and research value. Fault diagnosis and maintenance decision can not only reduce the maintenance cost of equipment, but also reduce the risk of accidents. Centrifugal compressor, as an important energy conversion device, has been widely used in the industrial field. Its working state and operational reliability have an important impact on the whole production system. As a kind of rotating machinery, the working condition of centrifugal compressor rotor has great influence on the stable operation of compressor. There are many causes of centrifugal compressor failure, such as machining error, medium corrosion, mechanical damage of material characteristics, excitation, flutter, surge, wear and so on, which are the main reasons that affect the normal operation of centrifugal compressor. In this paper, three typical faults of centrifugal compressor, such as oil film vortex, surge and rotor unbalance, are studied. The vibration characteristics of centrifugal compressor are studied by analyzing the radial vibration signal of compressor under the condition of fault. As the basis of fault diagnosis, the application of ant colony algorithm in fault diagnosis of centrifugal compressor is further studied. Ant colony algorithm is a heuristic bionic algorithm which has strong pattern recognition ability and can be used for data clustering analysis. In this paper, MATLAB is used to establish ant colony algorithm model, and simulation experiments are carried out to analyze the characteristics of measured vibration data. Through fault diagnosis of rolling bearings, the feasibility of ant colony algorithm in fault diagnosis of rotating machinery is verified. The diagnosis characteristic is studied, and the diagnosis method is further optimized and extended, and the actual fault condition of centrifugal compressor is analyzed and diagnosed. The result of diagnosis is compared with the actual situation, and the diagnosis effect is good and the program runs stably. The result is accurate.
【学位授予单位】：西安石油大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TQ051.21
【参考文献】
相关期刊论文 前9条
1 计晨;汪玉;杨莉;冯麟涵;;柴油机主要部件冲击响应时域分析[J];兵工学报;2011年04期
2 成正朝,牛大勇;组装式离心压缩机的特点及发展趋势[J];风机技术;2003年04期
3 丁克北;离心压缩机故障诊断研究现状及发展趋势[J];炼油与化工;2005年02期
4 白维峰;韩捷;董辛e，
本文编号：2422444
本文链接：https://www.wllwen.com/kejilunwen/huaxuehuagong/2422444.html
