基于噪声分析法的石油泵站机组故障自动监测技术研究

发布时间：2018-09-09 19:15

【摘要】：石油泵站单套机组由一部电机和一台离心泵构成,是石油集输生产过程的核心设备。为了避免因设备故障而导致安全生产事故,建立输油泵机组工作状态实时监测与故障诊断系统具有非常重要的意义。输油泵机组在运行状态下伴有振动和噪声,而这种特定的运行噪声含有丰富的设备状态信息,可以作为客观评估设备运行状态及故障诊断的依据。目前应用声学诊断技术进行设备故障诊断已经成为领域新的研究热点,尽管人们对于电机和离心泵各自运行噪声产生机理、典型故障成因及其引发噪声异变机理基本明晰,但是如何从整体运行噪声中准确的辨识各类典型故障尚需进一步研究。本课题提出基于噪声分析法的石油泵站机组故障自动监测技术研究,期待从技术层面提高输油泵机组运行状态评估的可靠性及故障判断的精准性。本课题融合虚拟仪器技术、声学检测与分析技术,以输油泵机组的组成结构及噪声特性参数为依据,以从石油泵站现场采集的机组运行噪声为分析样本,通过理论和试验相结合的方式进行系统研究。本文重点探索汽蚀故障所引发的噪声频谱异变,从而进一步确定汽蚀故障的检测方法与汽蚀发生初期判别算法。具体试验研究过程中针对输油泵故障在线实时监测过程中实现汽蚀故障初期状态的精准判别,需要依据具体机组汽蚀噪声特征确定汽蚀故障初期噪声比对分析样本。为此,提出基于特征频率分析的输油泵汽蚀故障初期噪声样本合成方法。本课题依据输油泵汽蚀故障噪声异变机理,在石油集输生产现场采集FS100-65-200机组运行噪声样本中捕捉汽蚀故障初期噪声频谱,按幅值递减方式从中提取N个单频信号,以此为基准基于LabVIEW实现汽蚀故障初期噪声样本合成。具体实现分三步:特征频率选择与标定、对应幅值设定、汽蚀故障初期特征噪声多频信息样本合成。试验证明,合成频率给定值调整步长为1Hz,给定频率与基准频率最大偏差0.32Hz,对应合成信号与基准信号幅值占比相对误差最大值1.90%,该方法针对不同型号输油泵机组具有可重复性和现实可操作性。
[Abstract]:The single unit of petroleum pumping station is composed of a motor and a centrifugal pump, which is the core equipment of petroleum gathering and transportation production process. In order to avoid accidents caused by equipment failure, it is of great significance to establish a real-time monitoring and fault diagnosis system for oil pump units. Oil pump unit is accompanied with vibration and noise in operation state, and this particular operating noise contains abundant information of equipment state, which can be used as the basis for objectively evaluating the equipment running state and fault diagnosis. At present, the application of acoustic diagnosis technology in equipment fault diagnosis has become a new research hotspot in the field, although the mechanism of noise generation, the cause of typical faults and the mechanism of noise aberration caused by motor and centrifugal pump are basically clear. However, how to identify the typical faults accurately from the whole running noise still needs further study. In this paper, the automatic fault monitoring technology of oil pumping station based on noise analysis method is proposed. It is expected to improve the reliability of operation state evaluation and the accuracy of fault judgment of oil pump unit from the technical level. The subject combines virtual instrument technology, acoustic detection and analysis technology, based on the structure of the oil pump unit and noise characteristic parameters, and takes the unit running noise collected from the oil pumping station as the analysis sample. Systematic research is carried out by combining theory with experiment. In this paper, the noise spectrum variation caused by cavitation fault is discussed, and the detection method of cavitation fault and the discrimination algorithm in the initial stage of cavitation are further determined. In the process of specific test and research, it is necessary to determine the initial noise comparison analysis sample according to the cavitation noise characteristics of specific units in order to accurately distinguish the initial state of cavitation failure in the process of on-line real-time monitoring of oil pump faults. Based on the characteristic frequency analysis, an initial noise sample synthesis method for oil pump cavitation fault is proposed. According to the noise variation mechanism of oil pump cavitation fault, the noise spectrum of the initial cavitation fault is captured in the sample of operating noise of FS100-65-200 unit collected from oil gathering and transportation production site, and N single frequency signals are extracted according to the mode of amplitude decrement. Based on LabVIEW, the initial noise sample synthesis of cavitation fault is realized. The realization is divided into three steps: characteristic frequency selection and calibration, corresponding amplitude setting, and multi-frequency sample synthesis of characteristic noise in the initial stage of cavitation failure. Experiments have proved that, The adjustment step of the synthetic frequency is 1 Hz, the maximum deviation between the given frequency and the reference frequency is 0.32 Hz, and the maximum relative error ratio of the synthetic signal to the reference signal is 1.90. This method has repeatability for different types of oil pump units. And practical maneuverability.
【学位授予单位】：东北石油大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TE977

【参考文献】