基于小波神经网络的开关磁阻发电机故障预测模型研究

摘要

开关磁阻发电机因其具有发电容量大、效率高、起动和发电组合容易等优点而拥有很大的应用前景和价值；首先利用MATLAB软件对开关磁阻发电机的电气故障进行了仿真,相电阻0.1Ω,导通角28°,关断角55°,转速7500rad/s,负载电阻114Ω,输出电容2000μF,得出故障下的特征数据,然后,鉴于小波神经网络预测算法结合了小波变换良好的时频局域化性质及传统神经网络自学习功能,建立了开关磁阻发电机基于小波神经网络的典型故障预测模型,仿真结果证明其良好的性能.%Swiched reluctance generator with the more high need to the power system of more-electric aircraft (MEA), which has the high value and apply foreground, the paper firstly finished the MATLAB emulation of electric fault of SRG, gaining the pressure signal, sec-ondly, because wavelet neural network combined the good time-frequency property of wavelet transformation and the studying ability, es-tablish the Fault forecast model of SRG based on wavelet neural network, the simulation result certificate the good function.