毕业设计论文基于神经网络异步电动机故障诊断.doc

摘 要 主要阐述了BP 神经网络在直流电动机故障诊断方面的应用。内容包括BP 神经网络的建立, 基于Matlab simulink的网络仿真三相异步电动机的运行状况直接影响到生产的正常进行，因此研究电机故障诊断技术，具有重大的理论意义和社会经济效益。 针对三相异步电动机的接地短路的外部故障，提出了一种基于BP神经网络的故障诊断方法,然后利用FFT 分析, 将振动信号的频谱分析作为神经网络的训练样本。对所采集异步电动机的定子转矩电流进行数据预处理与特征提取、归一化后，把这些特征参数作为神经网络的输入，经过学习训练，以判断系统状态，识别系统的故障。 通过选择足够的故障样本来训练神经网络, 将代表故障的信息输入训练好的神经网络后, 由输出结果就可以判断发生的故障种类。仿真和测试结果表明了该方法的有效性和正确性。 关键词:三相异步电动机;故障诊断;神经网络;BP算法 Abstract The application of BP neural network in the fault diagnosis of motor is explained. It contains setting up of the network and the network simulation based on Matlab simulink under the programming language environment .As the working status of the three-phase asynchronous motors directly impact on the daily order of production activity，it is very important to investigate the fault diagnosis techniques for the three—phase asynchronous motors. So it is of great theoretical and socio-economic benefits to study on electrical fault diagnosis technology． Aim at the faults of there-phase asynchronous motors such as ground fault, brings out one method of fault diagnosis based on BP neural network, then by FFT analysis, the frequency information of vibration is used as the training specimen of neural network. This method used characteristic information of asynchronous motor such as stator current finishes data preprocessing，feature extraction，and normalization. Then it uses these characteristic parameters as the inputs of the neura1 network，studies and trains，judges the state of system，and recognizes the fault of system. When symptoms that represent faults are input to the t rained neural network, the type of fault can be determined in the output of the neural network. The simulation and the test results point out its validity and correctness. Keywords: three-phase asynchronous motors; faults diagnosis; neural network; BP arithmetic 目 录 TOC \o 1-3 \h \z \u HYPERLINK \l _Toc295114897 第一章 绪 论 PAGEREF _Toc295114897 \h 1 HYPERLINK \l _Toc295114898 1.1 异步电动机工作原理及用途 PAGEREF _Toc295114