基于BP神经网络水泵电机的故障诊断

发布时间：2017-12-27 02:11

本文关键词：基于BP神经网络水泵电机的故障诊断　出处：《湖北工业大学》2012年硕士论文　论文类型：学位论文

【摘要】：水泵电机是供水系统中十分重要的设备。随着工业生产和行政部门对供水需求的提高使得水泵电机的运行时间和强度都超过了以往，水泵电机出现故障问题的概率明显增高了。目前由于生产成本和检测仪器的限制，多数情况下还是依靠专业人员的经验进行维护。这种方法对突发性和多样性类型的故障束手无策，，不得不停止运行进行修理，导致无法提供用水造成经济损失和社会负面影响。为了及时准确的预测和发现故障保证水泵电机的正常运行，避免水泵电机的因故障停机而造成的经济损失和社会负面影响，将BP神经网络理论应用于水泵电机的故障诊断就十分有研究意义。本文首先在分析了各种故障诊断方法的基础上对水泵电机的转子断条、定子绕组的匝间短路和轴承内外圈的这三种故障的产生机理作了全面的阐述，找出这三种故障对应的特征频率。然后采用了BP算法和网络模式，通过matlab软件选取不同参数进行讨论和训练仿真，分析仿真结果，找到性能最好的BP神经网络。文章最后针对BP神经网络自身的不足，将遗传算法与传统BP神经网络的算法相结合然后进行仿真，通过对结论的分析证实了遗传算法和BP算法的结合对BP神经网络的优化是行之有效的。
[Abstract]:The pump motor is a very important equipment in the water supply system. With the increase of industrial production and administrative departments' demand for water supply, the running time and intensity of pump motor are all higher than before. The probability of failure of pump motor is obviously increased. At present, due to the limitation of production costs and testing instruments, most of the cases still rely on the experience of professional personnel to maintain. This method is helpless for sudden and diversified failures. It has to stop operation and repair, resulting in failure to provide water and cause economic losses and social negative effects. In order to predict and find faults in time and accurately, ensure the normal operation of pump motor and avoid the economic losses and social negative effects caused by the failure of pump motor, it is of great significance to apply BP neural network theory to the fault diagnosis of water pump motor.
【学位授予单位】：湖北工业大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP183;TU991.35;TH165.3

【参考文献】