摘要
针对传统BP算法即梯度下降法,收敛速度慢、容易陷入局部极小值等缺点,提出了基于附加动量法和自适应学习速率的改进方法。将改进后的BP神经网络应用于汽轮发电机组的故障诊断中,获得了很好的效果,证明了改进方法的有效性。
Aiming at the defect of gradient-descending arithmetic, such as low-speed of constringency, easily falling into local minimum, an improved approach based on affixation momentum and self-adaptive learning rate is proposed. The results show the validity of the approach, which is applied in fault diagnosis of turbine-generator unit.
出处
《汽轮机技术》
北大核心
2006年第3期227-229,共3页
Turbine Technology
关键词
神经网络
故障诊断
汽轮发电机组
附加动量法
自适应学习速率
neural network
fault diagnosis
turbine-generator unit
affixation momentum
serf.adaptive learning rate