摘要
结合人工神经网络与智能容错控制,形成船舶柴油发电机转速神经网络容错控制。对由故障诊断后获取的特征值进行归一化处理,把经过处理的特征值作为神经网络的输入样本集,设计输出样本集,建立BP神经网络和Elman神经网络,用整理后的数据训练神经网络,使神经网络具有容错控制功能,并对神经网络模型进行仿真测试。仿真试验显示可以实现对船舶电力系统容错控制,保证船舶的安全运行。
Rotation speed control of marine diesel generator is built by the combination of artificial neural network ( ANN ) and intelligent fault tolerant control. The characteristic values obtained by fault diagnosis is treated together and the treated values are taken as input sample sets of neural network. The BP and ELMAN neural network are built by output of sample sets and trained with the treated data, so fault tolerant function is obtained. The simulation test is carried out for the model of neural network and the result shows that fault tolerant control of marine power system can ensure safe operation of ships.
出处
《防爆电机》
2009年第2期19-22,45,共5页
Explosion-proof Electric Machine
关键词
船舶电站
发电机
转速控制
神经网络
容错控制
Marine power station
generator
rotation speed control
neural network
fault tolerant control.