MATLAB在汽轮机振动故障诊断中的应用  

Application of MATLAB in Diagnosis of Turbine Vibration Failures

作　　者：程卫国[1] 傅志中[1] 陆文华[1] 钱安家[2] 

机构地区：[1]上海理工大学光学与电子信息工程学院,上海200093 [2]上海发电设备成套设计研究所,上海200240

出　　处：《动力工程》2005年第1期97-101,共5页Power Engineering

摘　　要：为了提高汽轮机故障诊断系统的诊断精度和自适应性,尝试了将人工神经网络引入汽轮机故障诊断的研究,通过对振动信号的分析,建立了汽轮机振动故障特征表及人工神经网络目标输出,利用MATLAB仿真软件的人工神经网络工具箱,设计构造了汽轮机振动故障诊断的人工神经网络,并对故障诊断人工神经网络进行了仿真训练和故障诊断。To improve the diagnosis accuracy and self-adaptability of steam turbine failure diagnosis systems, it is attempted to apply artificial neuro-networks (ANN) in research work concerning steam turbine fault diagnosis. Based on the analysis of vibration signals, tables of turbine vibration failure characteristics and ANN target outputs have been established, and a ANN composed, designed specially for diagnosing turbine vibration failures with the help of MATLAB-ANN tool kit, which has moreover been used for simulation training and failure diagnosis.

关 键 词：动力机械工程 汽轮机发电机组 故障诊断 MATLAB应用 人工神经网络 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]